This Week in Tech 1060 Transcript
Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.
Leo Laporte [00:00:00]:
It's time for TWiT this Week in Tech. Sammal. Sam, my car guy is here. We'll talk about cars spying on you. Mike Elgin the Gastro Nomad is here. We'll talk about AI with him. And from Windows Central, Daniel Rubino. We've got Black Friday sales data.
Leo Laporte [00:00:16]:
The statistics are interesting and they may not say what you think. They say the end of the Ford F150 and the Robox problem is a symptom of something much worse. All that and more coming up next on TWiT.
Sam Abuelsamid [00:00:33]:
Podcasts you love from people you trust.
Leo Laporte [00:00:37]:
This is TWiT. This is TWiT this Week in Tech. Episode 1060, recorded Sunday, November 30th, 2025. A shortage of shame. It's time for TWiT this Week at Tech, the show where we cover the week's tech news. I have combined a love, as I do every week, and with the help of Benito Gonzalez, who is on vacation this week, and Anthony Nielsen. Hello, Anthony. Filling in for Benito.
Leo Laporte [00:01:10]:
We always are looking. It's like a recipe. I remember Becky Worley told me this years ago. When you're putting together a twit, Becky said, you've got to have the meat, you got to have the veggies, and you have to have the spice. I don't know who's who on this show. Mike Elgin. You're pretty meaty. Mike Elgin is here from machinesociety, AI and gastronoma.net in the states for the holiday.
Mike Elgan [00:01:34]:
Right. I figured I'd try something different. And I'm here in California visiting with family.
Leo Laporte [00:01:39]:
Lovely.
Mike Elgan [00:01:40]:
And having a great time.
Leo Laporte [00:01:41]:
Good. Well, it's great to have you. We're welcome back to our shores. Where are you going next?
Mike Elgan [00:01:46]:
Well, we're going to Central America, and we're probably going to take a trip in December, but haven't quite decided where. So just. Just for fun.
Leo Laporte [00:01:55]:
You got itchy. You really itchy feet. I would be. If I were, you know, you and Amira, your wife. I would just say let's stay here for a month.
Mike Elgan [00:02:02]:
Yeah.
Leo Laporte [00:02:02]:
But nope. Nope.
Mike Elgan [00:02:03]:
The reason you say that is you have a house, you have a kitchen, you have all the things. Even when we're in California, we're still living out of suitcases, so you might as well. Why would we stay here?
Leo Laporte [00:02:15]:
Why would we stay here? It's suitcases all the way down. That's Daniel Rubino. He's also here from Windows Central. Great to see you. I forget. Where are you, Daniel?
Daniel Rubino [00:02:26]:
Worcester, Massachusetts.
Leo Laporte [00:02:27]:
That's right, Worcester. Lisa, yesterday Was reading place names. For some reason she said Worcestershire. Worcestershire.
Daniel Rubino [00:02:35]:
No one gets right.
Leo Laporte [00:02:36]:
Just ignore all the extra syllables in the middle.
Mike Elgan [00:02:39]:
Yes.
Daniel Rubino [00:02:40]:
Or we just call it the woo.
Sam Abuelsamid [00:02:42]:
The woo.
Leo Laporte [00:02:43]:
I like it. I grew up in Providence. So Worcester was just up the road a piece. Yes, I know Worcester well. And that in the middle. From Ypsilanti, Michigan. My friends, I give you Mr. Wheel Bearings himself, Samo Salmon.
Leo Laporte [00:02:56]:
Hello Sam.
Sam Abuelsamid [00:02:57]:
Hello Leo. How are you today?
Leo Laporte [00:02:59]:
Our car guy? I am. Well, whenever Sam's on, I, I always ask him what you driving this week.
Sam Abuelsamid [00:03:07]:
I have a Mercedes AMG G63 in the driveway which you know is the, the G Wagon is one of the most capable off road SUVs in the world.
Leo Laporte [00:03:18]:
And yet nobody buys a Mercedes to go off road.
Sam Abuelsamid [00:03:21]:
Well, well, and, and this, you know, some people do, but what they do is they buy 25 year old G Wagons.
Leo Laporte [00:03:26]:
Yeah.
Sam Abuelsamid [00:03:26]:
And then lift them and put proper wheels and tires because this one has 22 inch wheels with tires. You know, the, the only plover three locking differentials. The only place you can actually take it is driving down Rodeo Drive to Gucci.
Leo Laporte [00:03:41]:
You know, that's who buys.
Sam Abuelsamid [00:03:43]:
Yeah. In, in Beverly Hills, you know, Bel Air and in Miami beach. That's, that's where people drive these things.
Daniel Rubino [00:03:50]:
We were.
Sam Abuelsamid [00:03:50]:
Nobody drives new G Wagons off road.
Leo Laporte [00:03:53]:
Begonia, the new movie last night. And she drives a Mercedes off Jeep. It looks like a Land Rover. Lisa says that a Mercedes. I said yeah, yeah, that's, that's what.
Sam Abuelsamid [00:04:05]:
G, that's G Wagon.
Leo Laporte [00:04:06]:
Yeah, that's it.
Daniel Rubino [00:04:07]:
Yeah.
Leo Laporte [00:04:07]:
Those are, those are pretty cool looking.
Sam Abuelsamid [00:04:09]:
Yeah, I mean they were, they were originally designed in the late 70s by Mercedes at the request of the Shah of Iran. He wanted them for his, for his army, you know, instead of jeeps, you know, they wanted, you know, because Mercedes builds a lot of utility vehicles. They're big trucks and buses and stuff.
Leo Laporte [00:04:25]:
Yeah, yeah.
Sam Abuelsamid [00:04:26]:
And you know, so they wanted a really good durable off road utility vehicle. And so Mercedes developed the G Wagon and just as it was ready to go into production, you know, they had a little problem there in Tehran and he lost his job, couldn't pay for him anymore. Yeah. And so Mercedes decided to sell just. Well, you know, we got this. Let's just sell it to whoever wants it. And so they've been selling G Wagons ever since.
Leo Laporte [00:04:52]:
I think they're pretty cool looking. It's a little weird to see the Mercedes badge on something so utilitarian.
Mike Elgan [00:04:59]:
Yeah, I'm sorry. I was in the Cotswold about a month ago a Month and a half ago and met an elderly English gentleman who told me the most Texan story I've ever heard. He used to be in the industry of shipping oil equipment around the world. And some Texan oil owner of an oil company said, hey, I know you ship equipment in these things, but can you ship my Rolls Royce to Houston, you know, to Texas for me in this thing? He's like, what about customs?
Leo Laporte [00:05:25]:
What about.
Mike Elgan [00:05:26]:
He's like, don't worry about that. I'll take care of that. So the guy ships it to him. And then he later asked, how's the Rolls Royce going? He said, yeah, it's great. I converted it into a pickup truck. And he said, well, why would you convert a Rolls Royce into a pickup truck? He said, I'm kidding. I have the only Rolls Royce pickup in the world. This is like.
Leo Laporte [00:05:45]:
Anyway, I just thought it's like the mullet of trucks. It's like it's a luxury front and business in the back.
Sam Abuelsamid [00:05:52]:
The thing, the thing, the thing about Rolls Royce though is, I mean, you wouldn't have to convert it yourself. You just say, build me one. Yeah, build me one. Like this, you know, whatever color you want. You know, I mean, they do things like in the, they have these headliners with fiber optics in them that, you know, create a star field in the roof of the car. And if you want, they will, you know, you can give them any date and location and they will create the pattern of the sky on that, that night so you can, you know, have the, the star pattern of the night you were conceived or something. Lord, yeah. I mean, they will, they will literally build, you know, a blank check.
Sam Abuelsamid [00:06:35]:
They will build whatever you ask for.
Leo Laporte [00:06:37]:
And of course, that fabulous British hand done wiring is always exciting.
Sam Abuelsamid [00:06:42]:
You never know what could go wrong.
Mike Elgan [00:06:44]:
What could possibly, anything could happen.
Leo Laporte [00:06:46]:
Did any of you buy a Rolls Royce on Black Friday? This is, we are Thanksgiving post Thanksgiving Coma Friday. Black Friday Monday is Cyber Monday, which I don't think is meaningful anymore because that was all that was created in the day when you would wait to buy your stuff online to the. You went to work for the fast Internet on Monday. But now people have faster Internet at home.
Daniel Rubino [00:07:09]:
The whole thing has been diluted. All I've been doing is working that this weekend on our site.
Leo Laporte [00:07:13]:
Black Friday is all.
Mike Elgan [00:07:14]:
It starts in August.
Daniel Rubino [00:07:16]:
Well, Amazon started.
Sam Abuelsamid [00:07:17]:
It's been going for months.
Daniel Rubino [00:07:18]:
Amazon called Black Friday last Friday. And it's like they just cheat now. They just call whatever they want. It's cheating. Cyber Monday starts Saturday.
Sam Abuelsamid [00:07:27]:
It's all Made up. It was always all made, it's all diluted.
Daniel Rubino [00:07:30]:
It used to be like as corny as it was. It was like a thing, right. Friday morning people would go to the mall, line up and all this. And yeah, it was like an event. Now it's not an event, it's just watered down.
Sam Abuelsamid [00:07:44]:
I remember sitting outside best buy at 4:00 in the morning, waiting for the store to open.
Leo Laporte [00:07:48]:
Right.
Sam Abuelsamid [00:07:49]:
Yeah. And then, you know, getting a TV and a receiver or something. I can't remember what.
Mike Elgan [00:07:54]:
You know, I remember reading about those stories of people getting trampled.
Leo Laporte [00:07:58]:
Yeah.
Mike Elgan [00:07:59]:
Crushed by big screen tv, each other up and holy cow, those are the days.
Leo Laporte [00:08:03]:
Those are the days. Mike Paul Ferrat always tells a story of when his son was little. Of waiting outside of Target, I guess, to get the latest game.
Sam Abuelsamid [00:08:15]:
Call of Duty. The latest Call of Duty.
Leo Laporte [00:08:17]:
And it's like those. That's ancient history already, right? Everything's digital delivery. Well, according to spending reports, I think. Is it Adobe, whoever does these. Yeah, like Adobe Analytics. U.S. shoppers spent $11.8 billion on Friday online. Weirdly, I mean, I don't know.
Leo Laporte [00:08:40]:
I mean, I guess the economy is doing all right, up 9% from last year. This is still the biggest shopping day of the year. Yeah, AI powered shopping tools, a big player in all of this according to Adobe. And so this is the new thing. Let the AI do it.
Mike Elgan [00:09:01]:
Except they buried the lead on this. The sale of actual items is down 1% and the sales are up mainly because of inflation and tariffs. And so this, this BS that sales are up and the BS that AI is involved. I've seen no evidence that I had anything to do with this.
Daniel Rubino [00:09:22]:
Yeah, the AI tools, the AI tools are not accurate right now to say.
Leo Laporte [00:09:26]:
They want to improve.
Mike Elgan [00:09:27]:
People are using them. But what's the correlation between using AI and boosted sales? I mean, it's just everything's more expensive.
Leo Laporte [00:09:35]:
So this, this data is from Salesforce, which said order volumes were down 1%. Average selling prices up 7%. Consumers purchased fewer items at checkout with units per transaction falling 2% on a year over year basis. So really it's inflation.
Mike Elgan [00:09:55]:
People are buying less, everything costs more.
Sam Abuelsamid [00:09:57]:
That's the story.
Leo Laporte [00:09:59]:
Yeah. AI though increased 805% AI driven traffic to US retail sites eight times higher than last year according to Adobe.
Sam Abuelsamid [00:10:10]:
Yeah, but, but you know, keep in mind that when you're talking about going up 800% from a very low number.
Leo Laporte [00:10:17]:
Yeah.
Sam Abuelsamid [00:10:17]:
It's still not necessarily a very large number. Well, also when you starting from zero every percent is has a huge impact.
Leo Laporte [00:10:25]:
We didn't have Sparky and Rufus last year either. Sparky, for those of you who are not in the know, is Walmart's AI tool. Rufus is Amazon's AI tool. Have either any of you can ever clicked on Sparky or Rufus?
Daniel Rubino [00:10:41]:
No, no. I will say though, I mean these companies have done like, I like how Amazon has done some of this, right? So assuming you believe the customer reviews, which there are separate tools, so even gauge the accuracy of those, right. I do like the AI summarization use of that where they will summarize what customers are saying and then they'll pick out keywords about it and say, you know, which you can then click as hyperlinks and then learn more about that specific thing. So say if you're looking at a laptop and it'll say battery life and you click that and it'll have a negative. You'll see what people just say about battery life. So I think these are smart uses of this. But fundamentally all this is proportional to the cost of the item. So if something is like 10, 20 bucks, you'll, you may just use AI just to be like, here's the best thing.
Daniel Rubino [00:11:32]:
If you're going to spend $3,000 on something, you're going to ultimately want to see what a real human being is saying about it. And this is where I think the authenticity of this stuff kind of comes into effect. Where you want to watch someone with real photos, real video, real life experience with the product to sort of understand it better and sort of answer your questions. That's something AI I don't think will.
Leo Laporte [00:11:56]:
Ever figure out in the Amazon is interesting. Their current ad campaign doubles down on the reviews. Have you seen it? It's Benedict Cumberbatch. A real actor reading real reviews.
Sam Abuelsamid [00:12:07]:
Do we really know it's really him?
Leo Laporte [00:12:10]:
I didn't even think of that.
Sam Abuelsamid [00:12:12]:
Let me ask, I mean it could be a Sora cameo.
Leo Laporte [00:12:14]:
It could be. Yeah. Here I'm on Amazon. Happy Holiday Shopping. I'm Rufus, your shopping assistant. I can help you find gifts, compare products, check price history and more. I like the price history. That's a cool idea.
Leo Laporte [00:12:26]:
But my answers are powered by AI so I may not always get things right. So in honor of you, I'm going to Daniel, I'm going to ask what's the best price on a co pilot plus PC that I can get Searching the web. It does have a button for price history of items in my cart gathering products. Great news. What are the chances this will be on an Amazon deal.
Daniel Rubino [00:12:50]:
Oh yeah, exactly.
Leo Laporte [00:12:52]:
It says it's searching the web. Why bother? We know you're going to push me to Amazon.
Sam Abuelsamid [00:12:56]:
$259Amazon item.
Leo Laporte [00:13:00]:
Yeah, that's a good price. Is that. Should I buy that? Daniel, you're missing a lot of Copilot plus PC.
Daniel Rubino [00:13:13]:
We've been seeing a lot of big sales with the Qualcomm.
Leo Laporte [00:13:15]:
This one's intel actually but it's only got 8 gigs of RAM. I don't know how it could be a copilot plus PC. Don't they require more than that?
Daniel Rubino [00:13:24]:
Not necessarily, but the NPU if it does, the 40 tops.
Leo Laporte [00:13:31]:
This is pathetic.
Mike Elgan [00:13:32]:
There are two basic problems with Amazon's recommendation system. First of all, it has to be said that 25 years ago Amazon was the king of product recommendations with books, right? They were better than anyone. We'd never seen anything like it. They were able to tell you based on what you bought before what you would like to buy in the future. And that was based on what other people who bought those same books bought. And that was. That worked great. Nowadays there are two problems.
Mike Elgan [00:13:54]:
First of all, they don't really want you to buy the best thing. They want to favor the people who are playing the Amazon game. The vendors that are jumping through the hoops that Amazon wants them to profiting Amazon the most. So it's not really, they're not really looking out for you. This is one of the reasons why they have the recommended products and so on always score higher. They always have higher scores mysteriously than you know, you know than a product just like it that isn't hasn't been selected by Amazon to be a featured product. The second problem is when you order gifts especially their only metric for knowing if that gift is for you or not or some other person is whether you have it gift wrapped or whatever. But like they would know a lot more about what I want to buy for myself if they knew which of the things are for my granddaughter, which of the things are for my wife, which of the things are for somebody else, my kids, whatever.
Mike Elgan [00:14:49]:
I'm buying all this stuff. They couldn't possibly based on what I buy on Amazon profile me in terms of what I like to buy for myself.
Leo Laporte [00:14:57]:
Is that why the recommendations are so universally awful? They often recommend buying something I just bought, which exactly. Just kind of dopey. So but look, I just clicked. They said well do you want to know which has the best battery life? And Daniel, it gave me, it said the HP OmniBook 516 inch has up to 34 hours. That's a Snapdragon 450 bucks. Is that good? Is that good advice? I mean, if you're looking for battery life, that's a, that saves me going to Windows Central.
Daniel Rubino [00:15:27]:
Well, so maybe. Right, so the 34 hours, I think everybody on here probably realizes that that is, that seems like a lot.
Leo Laporte [00:15:34]:
That seems high.
Daniel Rubino [00:15:35]:
Yeah. So that's, that's what the manufacturer claims.
Leo Laporte [00:15:38]:
Right.
Daniel Rubino [00:15:38]:
So we review that laptop and we have the real world usage and DATs.
Leo Laporte [00:15:44]:
And you know, that's what you'd really want, isn't it? Not.
Daniel Rubino [00:15:46]:
That's what you want.
Leo Laporte [00:15:47]:
Right.
Daniel Rubino [00:15:47]:
So everybody knows there's a difference between the marketing language and that's where I think, you know, I know a lot of people in my job in this position hate AI. They're scared of it. For me it's an opportunity. It's just how you take advantage of it. But I think for what we do, because we actually use these products, right, That's. I always joke about Windows Central. It's not like we review a laptop and we all go back to our Macs. Like we actually do use Windows laptops.
Daniel Rubino [00:16:18]:
So like we actually do have real good knowledge about what's good out there and what we highly recommend. That's something that AI can't give you. Something that customer reviews are good, but they have their own problems obviously since there's astroturfing there. But that's where.
Leo Laporte [00:16:36]:
That's the problem with Amazon's AI. I would prefer to use something like Perplexity or Coggies Assistant or even ChatGPT and ask it because it's going to presumably give you more objective results, especially.
Mike Elgan [00:16:51]:
If you know what to, how to use it. If you, if you're good at prompt engineering, you know how to do role prompting, right. You can get really good advice especially go in and drill down and I've, I've, I've used different tools, AI tools to get product recommendations and I, I like to go in and based on what it tells me, I'm like, no, no, you're not really quite getting it. I really need something that's really light, lighter, light, that's the main thing. And you just go back and forth, forth a bit and you could get really good advice. And once you have that advice, then you can go to the E commerce sites and figure out where to buy.
Leo Laporte [00:17:23]:
It, I think so. For instance, Cogi told me it's important to remember real world battery life is highly subjective, depends on specific tasks. It actually quotes reviews and then talks about Benchmarks. There's Paul Thurat's site coming up on top, Forbes, Reddit, Tom's Guide. I'm sure Windows Central pop up here. That's. See, I think that's more useful.
Mike Elgan [00:17:49]:
Full disclosure, my son works at cogi, but yeah, Kagi's great.
Leo Laporte [00:17:52]:
I really love, you know, I pay for, I pay for a COGI Pro subscription and I prefer. I've actually started using that instead of Perplexity. I do like the idea of the AI orchestrators, the tools like Coggies Assistant or Perplexity that pull in from multiple AIs.
Mike Elgan [00:18:10]:
Yes.
Leo Laporte [00:18:10]:
And match search. Although now with ChatGPT5, they do, the search is built in.
Mike Elgan [00:18:16]:
But, but the real, the real difference with Kagi is that there's no ulterior motive. You pay for it. That's their business model that you pay them and then they're helping you.
Leo Laporte [00:18:25]:
Right.
Mike Elgan [00:18:26]:
The opposite of what I was saying about Amazon. Amazon's trying to help Amazon.
Leo Laporte [00:18:29]:
Right?
Daniel Rubino [00:18:30]:
Yeah.
Mike Elgan [00:18:30]:
And use you to help Amazon. Whereas Kagi, because you pay for it, is really trying to help you.
Leo Laporte [00:18:35]:
Yeah.
Daniel Rubino [00:18:36]:
And you know, getting into that a little bit more specifically, like Best Buy, I think has done a really good job. You know, Best Buy, if it was a different world, would be out of business at this point. But I think they've done a really good job in trying to pivot and combat Amazon. And so they often offer way better prices than Amazon on a lot of products that we endorse. So. And for some people, going to Best Buy is one, you can just go down the street, pick up your item that day and if you have a problem, you can return it. Talk to a human being. I'm not saying Amazon's like bad for returns, they're actually quite good.
Daniel Rubino [00:19:10]:
But people still like having that ability. And so that's the problem with those AI engines. Right.
Mike Elgan [00:19:15]:
Yeah.
Daniel Rubino [00:19:16]:
Amazon one is just going to tell you the Amazon laptops, but that laptop could be a better deal. Different configuration, doorbuster thing at Best Buy or Target or even Walmart. Right. So yeah, you do need to have the, an overarching AI almost to figure all that stuff out. But then fundamentally you're going to want to come back to a human being and people have actually used these things.
Leo Laporte [00:19:38]:
So we don't have the traditional brick and mortar store numbers yet. We have the. Obviously, because that's a little slower than online, but it does look like initial forecasts are people preferred online to shopping in brick and mortar. I guess that one. That's not really surprising. Yeah.
Sam Abuelsamid [00:19:57]:
I mean yeah, that's, that's a trend that has been going on for, yeah. Decade, you know, a couple of decades now that, that shift towards online shopping, especially Next Global Manager Black Friday, right.
Leo Laporte [00:20:08]:
Retail Next Global Manager of Advanced Analytics. According to this is a story from Forbes said the story isn't just that shoppers stayed home. It's that they're changing how and when they shop. We're seeing a consumer who's still spending but doing it with surgical precision. They're waiting for the right price, stretching purchases across a longer promo window and walking into stores with far narrower mission than we've seen in past holiday seasons. That seems like that's Internet driven, search driven, AI driven. People are smart.
Sam Abuelsamid [00:20:37]:
Even when I go to Best Buy, you know, I, I do all the research online and I figure out exactly what it is. Yeah, I want to buy, you know, and I walk in the door and actually, I mean, a lot of the times I'll just do the order online and just have it. Yeah. When it's, when it's ready, I go pick it up. But even if I don't do that, you know, the only time I spend in the store is however long it takes me to actually find that item in that particular store and then buy it and then get out. I don't, I don't browse in the store anymore.
Mike Elgan [00:21:08]:
You know, it's funny, it's funny because we bought some furniture lately for one of my sons and we went up to customize, do all this kind of stuff, and the person in the store who was helping us was just using their website.
Leo Laporte [00:21:23]:
I hate that. It's like going to a bookstore and they don't have a book. And they said, well, we can order that for you and have it five days, Right. I can order it for me and have it tomorrow, right?
Daniel Rubino [00:21:32]:
Yeah.
Leo Laporte [00:21:33]:
This is what's killing brick and mortar, I guess.
Daniel Rubino [00:21:36]:
Even though, even with Best Buy, I was just gonna say like, with Best Buy, the stores are often pretty empty because just what we're talking about. Right. A lot of people order online, then they go and pick it up, but you're only in there for a couple of minutes.
Mike Elgan [00:21:48]:
This is the advantage of stores that are all purpose stores instead of just electronic stores. For example, my wife and I were going to buy AirPods Pro 3. We're going to buy it online. We hadn't gotten around to it, but we went to Costco for some other thing and there they had the same price as online, plus they added AppleCare for free. So we're like, yeah, what the heck, throw it in the cart. Right. So Walmart and Costco and these sort of general purpose stores are more likely to capture people who would otherwise buy online. Whereas with Best Buy you're not going to go there unless you're intending to buy a specific, you know, electronics product.
Leo Laporte [00:22:23]:
Yeah.
Daniel Rubino [00:22:23]:
Costco also does exclusive configurations for a lot of laptops, which is pretty interesting.
Leo Laporte [00:22:29]:
Smart people like Costco and the electronics because they have kinds of things. Yeah, very generous. They use their buying power to get very generous return terms for consumers. So you can. It's easy to return stuff and they have a longer window than typical.
Mike Elgan [00:22:44]:
It's more than that, actually. We actually bought this huge, huge heavy table with a full set of chairs that was delivered and they put it together and then there was a problem with one of the chairs and they came to the house, disassembled it and carried it away as the return at no charge. Wow, that's unbelievable. That really makes you want to buy stuff.
Leo Laporte [00:23:06]:
They're smart. That gets people back in the stores. It says apparently, according to Adobe analytics, that Gen Z did in fact go into the stores. For some reason they like them all. But maybe it's what you were saying, it's one stop shopping, right? That they can go to one area, get their shopping done.
Mike Elgan [00:23:24]:
It's a novelty to them.
Leo Laporte [00:23:28]:
Look, they have many stores in one place. That's a miracle.
Sam Abuelsamid [00:23:33]:
Well, if you can actually find a mall that has many stores.
Leo Laporte [00:23:36]:
Well, that's right. Boy, most of the malls I know of have lost their anchor tenants. You know, Sears and Macy's and so forth are gone and they're like Marlboro.
Daniel Rubino [00:23:46]:
It was just announced that the Apple store is pulling out of Marlboro Mall at the Salomon Pond and they're going to move to a different mall.
Mike Elgan [00:23:55]:
Yeah, that's devastating to the mall because Apple is unique in. In mall retail. Basically they gauge the quality of a store and which is an anchor, which isn't an anchor. Other metrics based on the. The daily sales per square foot and you could take the second highest sales per square foot and Apple is 20 times higher.
Leo Laporte [00:24:20]:
Yeah. Better than a Tiffany.
Mike Elgan [00:24:23]:
Far more.
Leo Laporte [00:24:24]:
Yeah, it's very, it's very small.
Daniel Rubino [00:24:27]:
To the Salt Lake Paul mall to shut down, unfortunately.
Leo Laporte [00:24:30]:
Yeah, That's a big deal. Yeah, yeah. Fewer super big discounts this year. Absent from this year's promotions were last year's pervasive 60% off signs as retailers try to save price and mitigate the impact of tariffs on their margins. The season is rarely one on Black Friday and Cyber Monday retailers continue to have to entice customers. Yeah, I haven't done my Christmas shopping yet.
Mike Elgan [00:24:56]:
I don't know.
Leo Laporte [00:24:56]:
Do I think people are smarter also about Black Friday? They understand that that's just. And this is what one of the analysts said. Just a data point. It's not the story. It's just one thing.
Daniel Rubino [00:25:07]:
A very late swing with Black Friday, meaning that Thursday and Friday were really soft and we didn't see a lot of traffic.
Leo Laporte [00:25:14]:
Interesting.
Daniel Rubino [00:25:15]:
It wasn't until Saturday afternoo we started to see traffic really pick up. And it's pretty high right now going into Monday. But it's definitely. It was a very different experience where most shopping wasn't happening on Friday. And we saw that.
Leo Laporte [00:25:30]:
Are you just covering PC sales or do you. What do you. What.
Daniel Rubino [00:25:33]:
Well, this is, this is across our company in tech.
Sam Abuelsamid [00:25:36]:
So.
Daniel Rubino [00:25:37]:
So this is Tom's hardware, Tom's guide.
Sam Abuelsamid [00:25:39]:
Yeah.
Daniel Rubino [00:25:40]:
Tech radar. Yeah, all those.
Leo Laporte [00:25:41]:
But it's tech tends to be tech.
Daniel Rubino [00:25:44]:
Yeah. Well, like I think like 70% of our revenue and our and future is tech.
Leo Laporte [00:25:50]:
Tech. Yeah. Well, it's not a bad place to be right now.
Daniel Rubino [00:25:54]:
No, no. I mean it's a lot of. It's going around, right?
Leo Laporte [00:25:57]:
Yeah, it's doing quite well. All right, let's take a break. We will talk about how well AI is doing, especially for the guy, one of the hosts of the all in podcast. New York Times. Little name drop for all in the show run by our friend Jason Calacanis in just a moment. You're watching this Week in Tech with Windows Central Editor in Chief Daniel Rubino. I didn't mention that at the beginning of the show. I'll mention it now.
Leo Laporte [00:26:22]:
It's nice to have you from Worcester from Ypsilanti, it's the host of the Wheel Bearings podcast and automotive analyst at Telemetry, Samo Balsamit and Mike Elgin, our gastronomic nomad. Gastronomic nomad. Our gastronomad.net his AI blog is machinesociety. AI we'll talk about AI. There's a lot of AI news in just a little bit. I was actually bemoaning this to Lisa when I was going through stories this week because we have a show that's dedicated to AI intelligent machines with Jeff Jarvis and Paris Martineau on Wednesdays. And as I go through stories, I. I'm saying this one, allocating them to Twitter and it was all AI stories.
Leo Laporte [00:27:06]:
I don't have a long enough show on Wednesday to cover all these stories. So I'm I'm going to move some AI into this show. I hope you don't mind. There's a there. This is one of those areas in tech and I can't think of another time that it's been like this. Maybe you guys, maybe the Internet was like this, where it's just dominates, not just tech, media, but it dominates the market, dominates everything right now. It's the big story of the year, probably the decade. Who knows, maybe even in the century.
Leo Laporte [00:27:36]:
We'll get to that in just a little bit, but first word from our sponsor and we're talking about AI invention. AI is supposed to make things easier for you and your teams, but for a lot of teams, it's made the job worse, it's made the job harder. But that's where Vention comes in. Their 20 plus years of global engineering expertise can make the difference between an AI bust and an AI boom. They build AI enabled engineering teams, but not just, you know, let's put together a team. They want to make software development better, faster, cleaner and, yeah, even calmer. And they can do it. Clients typically see at least a 15% boost in efficiency, not through hype, but through real engineering discipline.
Leo Laporte [00:28:25]:
Vention's been doing this for a long time. They know how to do it right. The other thing you might want to look at Vention for is their incredible AI workshops. These are really fun interactive workshops for you and your team, designed to help you focus, to take the team, take those ideas and come up with a plan to help your team find practical and safe ways to use AI, cost, delivery and qa. It's a great way to start with Vention and test their expertise, see if these guys are right for you. Whether you're a cto, a tech lead or a product owner, these workshops help you so much you don't have to spend weeks figuring out tools and architectures. What model do you use? I mean, that's the perennial question. But Vention can help you steer your way in an interactive process to assess your AI readiness, to clarify your goals, and then to outline the steps to get you there without the headaches.
Leo Laporte [00:29:22]:
And then if you need help on the engineering front, they're there too. Their teams are ready to jump in, whether it's as a development partner or a consulting partner. Whatever role you need, they're there for you. This is the most reliable step to take after your poc. Let's say you've built a promising prototype in Lovable and it's running. It does well in tests. But now this is when the challenge happens. What's next? Do you open a dozen AI specific roles just to keep moving? Do you bring in a partner who's done this across industries? Someone with the expertise, the experience, Someone who can expand your idea into a full scale product without disrupting your existing systems or getting in the way of your team? Hey, that sounds pretty good, doesn't it? That's Vention.
Leo Laporte [00:30:08]:
Vention is real people with real expertise and real results. Learn more@ventionteams.com and see how your team can build smarter, faster and with a lot more peace of mind. Or as I said, get started with your AI workshop today. Ventionteams.com TWIT that's V E n t I o n like invention without the in Vention. V e n t I o n teams. Thank you Vention for your support of this week in Tech. We appreciate it. You support us when you go to that website.
Leo Laporte [00:30:42]:
Ventionteams.com TWIT I woke up this morning New York Times big story about David Sachs, who is of course the AI and crypto czar in the Trump White House. You know, we were talking before the show about it's kind of hard to know what ax the times is grinding at any particular time. But Sachs has definitely been very instrumental in the White House's AI action plan. Apparently he was the guy who said no state regulation of AI. We've got to find a way to stop that. Trump was considering an executive order. The last we heard is they're going to put it in the National Defense Authorization act, which is a great way to, you know, put little earmark pieces of legislation in because it's always passed, right? And very seldom is it edited. So it may well be that when the NDAA is passed that we suddenly have a federal law banning state laws against AI.
Leo Laporte [00:31:44]:
That includes California, Illinois, a lot of states trying to regulate AI. But Sachs represents the industry. He has this kind of stunned me. Seven hundred and some 708 tech investments, including 449 stakes in companies with ties to AI that could be added aided directly or indirectly by his policies. According to the New York Times, nearly conflict of interest there 450 investments. His public filings designate 438 of his tech investments as software or hardware companies, even though the firms promote themselves as AI enterprises, offer AI services or have AI in the names. And then nice little plug for the podcast AI Podcast all in podcast, which I don't think was intentionally AI, but I think it's become AI. He certainly helped the all in Podcast become number one.
Leo Laporte [00:32:45]:
I guess there's nothing really to say about it. I'm just jealous. I thought, I wish I had a position in the White House that I could promote our podcast with. Yeah, maybe not.
Mike Elgan [00:32:54]:
I mean, I think in general, it's theoretically a good idea to have national policy on something like this rather than state policy. And let's be honest, when we're talking about state policy, we're talking about California. All the AI companies are in California, most of them, anyway. And California is a little bit more like Europe in terms of its regulatory outlook. They tend to be more, you know, consumer protectionist and that sort of thing. And, and I don't particularly trust this administration. I mean, the reality is, if we're being honest, is that it's, you know, with the Trump administration, it's kiss the ring and win some prizes.
Leo Laporte [00:33:32]:
Right.
Mike Elgan [00:33:32]:
He tends to look at people, individual people who own things, and if they play ball with him, he plays ball with them.
Sam Abuelsamid [00:33:42]:
There's also the bring enough gold bars or trinkets.
Mike Elgan [00:33:45]:
Exactly. Or a CR down, maybe, or a jet.
Leo Laporte [00:33:47]:
But separate from that, this is. There's an interesting dichotomy here. On the one hand, I understand the AI company's points of view that you don't want a patchwork quilt of 50 states regulations to wend your way through. If there were a single federal regulation that they could adhere to that they understood that would be.
Sam Abuelsamid [00:34:08]:
And in ideal circumstances, that would be absolutely true. Ideally, you want a single federal regulation, but the problem is we, we live in a time where we have a regime in Washington that is fundamentally opposed to any regulation on any industry on which they could be making money. And so, you know, in the absence of federal regulation, we still need something. And if the feds aren't going to do it, we, unfortunately, we have to rely on the states to do it. And, you know, this applies in so many areas. Not, not just AI, but, you know, in so many other areas.
Leo Laporte [00:34:44]:
I'm going to confess to being an AI fan and AI accelerationist. I'd like to, you know, in the early days of the Internet, the, the policy of the government was hands off. This is a nascent technology. Let's see what happens before we regulate the baby before it gets to grow up. So in a way, I'm. I don't know, I might be in favor of a minimal AI regulation.
Mike Elgan [00:35:10]:
Okay.
Leo Laporte [00:35:10]:
Yeah, but they're, they're. Tell me why that's a bad idea.
Mike Elgan [00:35:13]:
Well, there are two issues we're talking about. We're talking about the Degree of regulation. We're talking about who does the regulating, whether it's states or the federal government. I like the way Europe does a lot of things because Europe is essentially what the United States was supposed to be, a bunch of independent states with an overarching regulatory framework. But oftentimes what they do is they issue guidelines. And so they say, here's what you ought to do. And most of the states go, okay, we'll do that. Right? So they do all the thinking, they do all the think tanking, they do all that stuff.
Mike Elgan [00:35:41]:
They issue guidelines. And then there can be tweaks. I mean, this is. We're talking about companies who are offering a product that can individually tailor for every single person on earth, an entire agent assistant or whatever. And they can't. They can't do with 50 states. I mean, they can handle the different regulations based on states. It should be easier for AI than any other company.
Mike Elgan [00:36:06]:
Regarding whether we start, you know, this is the old thing from the social network in terms of how much you regulate. We don't even know what it is yet. Why would we start, you know, messing with it, all that kind of stuff. I agree with you. I think we need to find out what happens while mitigating the harms to children. There's some clear harms that have emerged. We need to protect the public from those harms. But other than that, we.
Mike Elgan [00:36:30]:
I think we really should find out where it goes before we start suppressing it.
Leo Laporte [00:36:35]:
So David Sack's response to the New York Times is, I'm suing your asses. He says it was a nothing burger story that they've been working on for five months and didn't find any smoking gun. But he says as it became clear. This is his tweet, as it became clear the New York Times wasn't interested in writing a fair story. I hired the law firm Claire Locke, which specializes in defamation law.
Daniel Rubino [00:37:02]:
And he publishes Fragile Egos.
Leo Laporte [00:37:04]:
No kidding. No kidding.
Daniel Rubino [00:37:06]:
Just like the most sensitive snowflakes, he.
Leo Laporte [00:37:09]:
Calls it the New York Times hoax factory. Now, even now, maybe he'll sue Stephen Bannon, who has been, of course, a White House friend and foe. Bannon, in the New York Times article, is quoted as saying that Sachs is a quintessential example of ethical conflicts in an administration where, quote, the Tech Bros. Are out of control. They're leading the White House down the road to perdition with this ascendant technocratic oligarchy. That's Steve Bannon.
Daniel Rubino [00:37:41]:
I know it's weird when you're like.
Leo Laporte [00:37:43]:
I don't understand it. Yeah, I don't understand. And then this is the week the White House announced the ominously named Genesis Mission, which is basically. I mean, if I were to encapsulate it in a sentence, the full speed ahead on AI, let nothing get in our way. We've got to beat China mission.
Daniel Rubino [00:38:04]:
That's all this is about. It's China. This is a conundrum we're in because everybody's like, AI is going to destroy us. And then where everybody's like, well, maybe we should regulate it. And then. And then all the states rights people all of a sudden, no, you can't. You can't. No, states, right, don't apply here.
Daniel Rubino [00:38:20]:
It only applies to things that, like, we don't like. And then it's like, well, we can't regulate because China won't regulate. And since they're not going to. If they get ahead, they're going to own this market.
Mike Elgan [00:38:31]:
We don't want China's AI to destroy us. We want our AI.
Daniel Rubino [00:38:35]:
Yeah.
Leo Laporte [00:38:36]:
Red, white, blue AI destroy us.
Daniel Rubino [00:38:38]:
So we just let it fly. And this is like, why I'm, you know, it's like, I don't.
Leo Laporte [00:38:43]:
It's hard to find moral clarity here. For instance, the Genesis mission, which really sounds like the blueprint for Skynet, is in fact doing something, I think, very good, which is releasing the Department of all. The Department of Energy has all of these great labs, like the Lawrence Livermore Labs, all over the place. And these labs have a lot of research, a lot of work, not just on weapons, on everything, on materials science and all sorts of stuff that we paid for, but they've been kept private because, I guess presuming that enemy countries will use it against us or something, this releases that information to AI companies for training.
Daniel Rubino [00:39:26]:
That's alien technology, small UFOs.
Leo Laporte [00:39:29]:
This guarantees access to appropriate data sets, including proprietary, federally curated and open scientific data sets, in addition to synthetic data generated through DOE computing resources for use by private AI companies for their training. I think that's. I don't know. I think that's a good thing. This is why it's so hard to say.
Mike Elgan [00:39:58]:
I'm not sure it comes down to.
Leo Laporte [00:39:59]:
What your attitude towards AI is. If you think it's threatening to humanity, then of course you wouldn't want this. If you think it could be the most important invention humanity has ever come up with and could be transformational, then maybe you do want.
Sam Abuelsamid [00:40:12]:
You know, I don't. I don't think AI in and of itself is necessarily threatening to humanity. You know, I don't think AI is suddenly going to become sentient anytime in the foreseeable future and take over, you know, like a skynet type of thing. The problem is as AI becomes more sophisticated, you know it's going to be abused by people. It's humans that are the threat to humanity.
Leo Laporte [00:40:36]:
That's always been the case.
Sam Abuelsamid [00:40:38]:
Yeah, just, just as with any other technology. And the more powerful, the more capable AI becomes, the more ways there are for humans to abuse it and, and, and do things to society and to civilization and, and to the planet. You know, I mean, not the least of which is the amount of energy and water that is used for all these data centers.
Daniel Rubino [00:41:04]:
I mean, Palantir and law enforcement agencies with AI is just a recipe for the surveillance state that we were told.
Sam Abuelsamid [00:41:12]:
I don't, I don't want Alex Carp, you know, sticking his fingers into, into all of our information.
Daniel Rubino [00:41:19]:
Ironically, it's about money.
Leo Laporte [00:41:21]:
Ironically, though, remember Anduril, which is Palmer Lucky's company? Apparently Wall Street Journal is saying Anduril is stumbling in its tests with the military. In combat, we do fail a lot, says Palmer Lucky. Defense startup Andriel hits setbacks with weapons tech. And this is part of the problem too is you invest in all of this snake oil in many cases, and it ends up being not so hot. There's Palmer Lucky in his Hawaiian shirt. He, of course, you may remember the name, was the creator of Oculus and the Oculus Rift, which he sold to Meta and then left Meta under kind of a cloud.
Daniel Rubino [00:42:11]:
I will say he was smart for pivoting towards the fence because you get those contracts and you end up making, I mean, again, Palantir, you make a lot of money and those contracts stay locked in for a very long time. And there's not a lot of competition. I mean, in the defense industry, there are some really big players, obviously, but because his company is so much smaller that they can just innovate and fail faster than the big ones can. So he gets stuff done, I think a lot quicker now, whether or not it'll be successful.
Leo Laporte [00:42:41]:
Says Andrew's only real battleful experience in Ukraine has been marred by problems. In fact, the drones they made for Ukraine were so problematic, the Ukraine military stopped using them last year and haven't fielded them since. So anyway, but that's, that's, you know, it's a new tech. This is AI. It's a new technology.
Mike Elgan [00:43:01]:
Yeah, supporting your point, Leo. We, we are hyper focused on the chat bot dimension, which is not the most interesting. And. No. And Secondarily on the military. But actually if you list the things AI is being used for, those are tiny, tiny parts of that list. It's being used for medical diagnostics, for emergency response. It's being used to, to power prosthetics for, for, for disabled people.
Mike Elgan [00:43:27]:
It's there, it goes on and it's remarkable.
Leo Laporte [00:43:30]:
I think the promise is incredible.
Mike Elgan [00:43:32]:
Yes. And climate and climate, dealing with climate crisis.
Leo Laporte [00:43:35]:
So yeah, so many things material.
Daniel Rubino [00:43:38]:
Can AI predict when it's going to take over and destroy us?
Mike Elgan [00:43:41]:
Yeah, that'd be great.
Sam Abuelsamid [00:43:44]:
Can I predict when it's going to destroy our climate?
Daniel Rubino [00:43:47]:
Right.
Leo Laporte [00:43:48]:
Well, okay.
Daniel Rubino [00:43:49]:
And then fix it because it's smart enough to be able to fix it.
Leo Laporte [00:43:52]:
Well, yeah. And, and I also have to point out that yes, AI certainly is, is using a lot of energy, but so is a lot of other things. So are many of the other things that we are doing in our lives. And in fact, in some ways I think AI is driving, because it needs so much energy, is driving a move towards a better renewable energy.
Sam Abuelsamid [00:44:17]:
No, no, natural, they're just burning more natural gas and coal. I know there is coal fired. They're keeping ancient coal fired power plants running.
Leo Laporte [00:44:26]:
But get ready for this from Ars Technico. Over the course 2025, solar's growth is almost enough to offset the rising energy use.
Sam Abuelsamid [00:44:36]:
That's mostly in China.
Leo Laporte [00:44:39]:
No, in the US according to.
Sam Abuelsamid [00:44:42]:
Yeah, well that, I think, you know, that's, that's partly because that's a, that's a temporary glitch because I think everybody was scrambling this year to get stuff implemented, get stuff deployed before all of the tax breaks go away.
Mike Elgan [00:44:55]:
But also it's skewed in favor of what's happening in California. California. The amount of renewable, especially solar power that's, that's come online in California the last four or five years is incredible.
Leo Laporte [00:45:06]:
So, and so the first nine months of 2025, electricity demand has gone, has risen by a relatively speaking slow 2.3%. It isn't the through the roof nightmare that people are describing. And solar has made a big difference in that covering that rise. There's also a drop in natural gas usage, although coal usage is up by 23%. So this is the pie chart for where our electricity is coming from right now. Natural gas is 41%, 18% nuclear. That's the other thing. There's been a lot of work on modular nuclear plants, on natrium sodium cooled nuclear plants, on smaller nuclear plants.
Sam Abuelsamid [00:45:55]:
And how many of those have been deployed so far? None.
Leo Laporte [00:45:57]:
We work and it takes a while Sam, it's not overnight. You got to. Takes a while to build these. And yes, sure, we recommissioned Three Mile Island. What could possibly go wrong with that? But I don't think it's 100% a negative story. I think it's.
Sam Abuelsamid [00:46:16]:
I think, I think when we are measuring the power of data centers not by the actual how much compute they can do, but how much power they're consuming. Like talking about 10 gigawatt data centers, that's a problem.
Leo Laporte [00:46:30]:
That tells you something.
Sam Abuelsamid [00:46:31]:
That's a fundamental problem.
Daniel Rubino [00:46:32]:
Yeah, but as the chips and the GPUs and everything, the technology improves, they'll become more efficient over time. You know, I mean, I feel like a lot of the GPUs that even Nvidia using right now is mostly almost a stopgap to get these things out to market in such fast numbers that they can meet demand, but they're not necessarily the best at what they can be. And I think as the market is absolutely true, you know, we'll, we'll get better with that stuff.
Sam Abuelsamid [00:47:01]:
Yeah, do this by brute force. They're not, they're far, they're far from the most efficient way to do this kind of compute.
Mike Elgan [00:47:09]:
I mean, most of the energy is air conditioning, stuff like that. So dealing with heat is a big problem. But I, One of the, one of the unexpected ways, or maybe it is expected ways, that AI is increasing energy use is that people are reading about how much energy is required for these data centers and saying, you know, I used to unplug my nightlight in the middle of the night to say the environment. And these guys are, you know, so what the hell, I'm just going to, you know, just leave everything on all the time.
Daniel Rubino [00:47:36]:
Well, it was an open. I had a blog post talking about how every time someone would say thank you in a chat, it would like use like 10,000 gallons of water. Like, ridiculous number, you know?
Mike Elgan [00:47:47]:
Yeah. People read that stuff and they're like, what am I, Why am I, you know, sorting my recycling when this is what's happening.
Leo Laporte [00:47:55]:
One of the, one of the plans is to put these data centers, these AI data centers in space. This is a guy who has actually worked on us on this kind of thing. An article from Taranis, T A R A N I S. I'm not familiar with the publication, but I found this. It says data centers in space are a terrible, terrible, horrible. No good idea. There's a rush for AI companies to team up with space launch satellite companies to build data centers in space. TLDR it's not going to work.
Leo Laporte [00:48:24]:
He says he's a former NASA engineer and scientist with a PhD in space electronics. Worked for Google for a decade.
Mike Elgan [00:48:31]:
What does he know?
Leo Laporte [00:48:31]:
What does he know? He's worked on the bit of the cloud at Google responsible for deploying AI capacity. So he says, I'm quite well placed to have an opinion here. There are a lot of reasons why it's a bad idea. You know, one of the things people think is, oh, well, it's really cold out there, right?
Mike Elgan [00:48:50]:
Yeah.
Leo Laporte [00:48:50]:
But without an atmosphere to conduct the heat away, it's actually harder to cool stuff in space.
Daniel Rubino [00:48:56]:
Yeah, yeah. You don't instantly freeze in space. That is one of those movie tropes.
Leo Laporte [00:49:03]:
He's designed hardware to fly in space. So I think the guy knows a little bit. He says, we designed a camera system for space. Thermal management was front and center in the design process. It had to be because power is scarce in small spacecraft and thermal management has to be achieved. So he couldn't use heat pumps. Basically it turned down the power use because the cooling was so difficult. There's also radiation.
Leo Laporte [00:49:29]:
You gotta, you know, I mean, the chips that are sent out into space are often much less sophisticated than chips we use at home because of the cosmic rays and particles.
Sam Abuelsamid [00:49:38]:
Yeah, well, I mean, a great example of that, that is what's going on right now with Airbus. A couple of days ago, they had to issue a recall for 6,000 Airbus A320s, you know, which are the most common airliner in the world now because their electronic systems on many of them had gotten corrupted by recent solar storms.
Leo Laporte [00:49:58]:
Right.
Sam Abuelsamid [00:49:59]:
The software in there, and they had.
Leo Laporte [00:50:00]:
To downgrade the software.
Sam Abuelsamid [00:50:02]:
They had to reflash these things because is the, the just the. This and this is inside of our atmosphere where a lot of that stuff gets filtered out, you know, not out in space where there's nothing to filter it. So, you know, if, if we can't even protect our, our aircraft that are flying within our atmosphere, you know, the idea of having data centers in space, you know, does definitely seems like a bad idea.
Leo Laporte [00:50:28]:
Pigs in space. All right, I want to take a little break. And we come back, I will talk about China and its AI efforts, which, I mean, I'm, you know, I don't know if you noticed this. When I used cogi, I used a Chinese model with Kagi's assistant, you get to choose from a huge number. That's one of the reasons I, like, I'm sorry, I keep giving Kagi a plug. It's not because you're here Mike. I'm just a fan but I have been using glm which is a new model from a Chinese company that is Z AI which is quite good. But notice of the of the Kagi recommended models, three out of the four are Chinese.
Leo Laporte [00:51:10]:
The Chinese models are really good. We'll talk about that in just a little bit. They're also, by the way, one of the reasons Coggy likes them. They're among the least expensive models available out there. So that's good too. When you go to Claude, it gets a little pricey, little more expensive. You're watching this Week in tech. Mike, Elgin, Sammal, Samit and of course from Windows Central, great Daniel Rubino.
Leo Laporte [00:51:38]:
We are thrilled to have you here as well. Thank you for watching today. Our show brought to you by Deal. D E L. If this has happened to us, you know it's hard to do HR with people out of state or out of country. It makes it hard to hire. How many times has this happened to you? You find the perfect engineer, but they're overseas and then you realize that hiring them is going to be harder than getting your VPN to connect on the hotel wi fi. Harder than building a data center in space.
Leo Laporte [00:52:14]:
There's entities set up, there's local payroll laws, there's compliance hoops. It really can be a total nightmare. Deal fixes that one global AI powered platform that helps you hire onboard and pay anyone anywhere, fast and compliantly. No third party vendors, no duct taped integrations. Just one clean system for HR and for it and for payroll that scales with your team. You can spin up new hires in minutes, ship their devices, grant system access and run payroll in 100 plus countries. Think of it as DevOps for people operations global infrastructure that just works. The CEO of Airwallex says quote deal lets us hire and support top talent anywhere without slowing down.
Leo Laporte [00:53:04]:
Maybe that's why over 35,000 companies trust Deal. VisiT-E-E-L.com twit that's deal.com we thank them so much for their support of this Week in tech. So the Chinese models are getting better and better and even though we don't know that they are cheaper and cheaper to make because of course it's coming from China. It's their claims are that they are not spending as much money on power on GPUs from the Financial Times, China leapfrogs US and global market for open AI models. Oh and that's the other thing I like about these Chinese models. They are open source. Yeah. A study by the MIT and Hugging Face found the total share of downloads of new Chinese made open models rose to 17% in the past year, compared to 15.8% from American developers.
Leo Laporte [00:54:01]:
It's the first time Chinese groups have so Hugging Face, if you don't know, is a place you can go and download models. I use it all the time with LM Studio or you know, llama cpp. You could download the models and use them and run them locally. And I have to say for the most part I am downloading these open Chinese models, some of the ones I just showed you in Kagi also the one OpenAI model that is open, that's very good GPT OSS. But the Chinese models are free to download, they're free to modify. You can integrate them into your own stuff.
Mike Elgan [00:54:37]:
And this is the risk because obviously the Chinese government is interested in having control over information globally and by open sourcing these models, making them very inexpensive. And by the way, some reports say that they can produce these models at around 30% of the price as American models. We don't know if that's true or not, but they want to have information and they're already using this information. So for example, there's a report that came out recently that found that the deep seq R1LLM is, generates up to 50% more insecure code when prompted with politically sensitive inputs like Falun Gong, Uyghurs in Tibet.
Leo Laporte [00:55:18]:
In fact, it looked like it was intentionally producing insecure code for those things that were the enemy of the prc, of the Chinese.
Mike Elgan [00:55:25]:
Exactly. Well, well, you get what you pay for, Leo.
Leo Laporte [00:55:31]:
Well, well, well, of course you, you can download these models and get them uncensored and so forth. And, and I presume that that's post training that they're doing this kind of.
Mike Elgan [00:55:41]:
But it's a larger play, right? So it's a long term, the way the government thinks in China is, it's a long term big picture play where they, you know, they get everybody using these things. They have the power to update them. The state has major input into what happens at these companies. In fact, it's Chinese law that every Chinese company has to cooperate with Chinese intelligence.
Leo Laporte [00:56:03]:
Right.
Mike Elgan [00:56:04]:
And, and for leaders of those companies that don't comply off to jail with, you know, join the Uyghurs in the, in the, in the labor camps. But, so, so this is, this is a long term play and it's a real problem for the United States because, because in the United States we like big money, massive investment, expensive, you know, development of, of, of technologies and we're at a bit of a disadvantage for the models that the world uses.
Leo Laporte [00:56:34]:
Yeah. In fact, the Commerce Department has expanded its list of Chinese companies on the entity list. Some companies that we know well that are being widely used. In fact, we're watching with interest to see what happens to dji, the makers of those incredible drones.
Mike Elgan [00:56:52]:
I love those drones so much. I hope they don't get banned. I just need them.
Leo Laporte [00:56:56]:
But the problem is, you just stated the problem, that every Chinese company, by law, is kind of beholden to the Chinese government. So you could really say that every single product produced in China has that risk, right?
Mike Elgan [00:57:11]:
Yeah, I mean, it varies according to the type of product. So for example, routers and networking equipment. These are highly risky for, you know, and, and things.
Leo Laporte [00:57:22]:
Right.
Mike Elgan [00:57:23]:
Kind of. I mean, could you imagine anything that's got connectivity, connectivity, cameras, risk location capabilities, like, you know, the connections to phones. There's all kinds of thing problems with drones, unfortunately. They're, you know, toys and things like that. Who cares? But like, you know, these are, these are potentially usable. I mean, there, there's robot dogs, right? They, they found backdoors in Chinese robot dogs. You know, they could just activate them remotely, theoretically, and create a robot dog army in other countries controlled by Beijing.
Daniel Rubino [00:57:57]:
I mean, this is why they don't sell a lot of those Chinese EVs too, in the United States.
Mike Elgan [00:58:02]:
Well, and also there's a scandal.
Leo Laporte [00:58:05]:
So BYD, which is the biggest seller of EVs in China, is on this recently released from the Defense Department list of Chinese companies that are cooperating with or supporting the Chinese military. Alibaba, Baidu, and BYD. But Sam, everybody seems to agree that Chinese EVs are light years ahead of every other company outside of China. Is that right?
Sam Abuelsamid [00:58:28]:
Pretty much, yeah. I mean, they, they have moved so far in the last five years in terms of their software capabilities, their overall cost and efficiency, that. Yeah, they are as much as anything built anywhere. Yeah, yeah.
Leo Laporte [00:58:45]:
Alibaba is the largest E commerce company in China. Baidu specializes in AI and autonomous driving technology. And byd, am I right? They're the number one EV manufacturer in China? I think they are, Yeah.
Daniel Rubino [00:58:57]:
I think it was like, wasn't it six months ago they showed their car has AI that automatically adjusts the suspension in real time. And they showed it like going up a thing of stairs and the tires were adjusting in real time through AI. And so the people inside, it was a smooth ride for them. But the car was like going upstairs. I mean, that was like six months ago. They were showing that off very cool stuff.
Mike Elgan [00:59:20]:
But there was a scandal that started in Norway where they did tests underground where they could block all the signals and found that they were remote control capabilities in their Chinese buses that were used for public transportation. The ability to shut down the buses remotely. And so other, other countries like the UK started worrying about theirs. Israel has canceled all the Chinese bus contracts they had with the military and they're looking at, you know, for, for domestic use and. Which caused this, triggered this whole bigger thing. Well, like, what else, where else do they have back doors and kill switches and all this kind of stuff, you know. And you, you sound like a lunatic conspiracy theorist to suggest that, you know, the Chinese government is orchestrating this global distribution of electronic products that can be controlled remotely. But on the other hand, you're like, why is this stuff.
Sam Abuelsamid [01:00:09]:
To be fair though, every company has those kinds of capabilities. GM and OnStar in 2008 launched remote shutdown capabilities in OnStar for stolen vehicles. So if you reported your car stolen, they could reach out and make the car slow down and, you know, give the location to police if you didn't slow it down. Yeah. And that's not even from a manufacturer. That's. That's, you know, used car dealers putting that, you know, disabled, disabled systems immobilizers in their vehicles. You know, if you didn't pay your make your loan payments.
Mike Elgan [01:00:44]:
But the difference is that GM is not planning to invade Taiwan and, and.
Leo Laporte [01:00:48]:
And that it's not, thank God. Yeah, we talked about this a few weeks ago. We had Alex Stamos, brilliant security reacher researcher on Twitt, and he said Chinese are in our grid inextricably. We know they're in our telecommunications system inextricably. They're just sitting there. Why are they there? What are they up to? They're just waiting.
Mike Elgan [01:01:13]:
Why are they harvesting the personal, you know, harvesting the databases of government employees that have security clearances and building these. You know, you're just getting all this stuff for, you know, obviously they think big picture. They love tons and tons of data. They want to be prepared for future conflicts. They kind of learn from the US in the Iraq invasion.
Leo Laporte [01:01:33]:
Here's my argument. I think this is defensive, not offensive. And I think the best way to handle this is to make China already is huge financial ties to the United States just through loans. You know, they own half of New York City.
Daniel Rubino [01:01:46]:
That's the argument. Right. The closer our economies are.
Leo Laporte [01:01:50]:
Yeah. The safer way forward to peace is to work with them, is to cooperate with them, is to trade with them. Economically dependent on us. Yeah. Not to shut down and, and, and increase the paranoia around them.
Daniel Rubino [01:02:06]:
It's an economic mad. Exactly.
Sam Abuelsamid [01:02:10]:
That's right. The problem, though, is that under President, President Xi over the last decade, I think, I think it started in 2015, they started a policy of not so much decoupling from the west, but making the, the, the local Chinese economy less dependent on the West.
Leo Laporte [01:02:29]:
Again, defensively, I would do that if I were them too. Yeah.
Sam Abuelsamid [01:02:33]:
But, you know, so, you know, they, they encouraged people to buy Chinese brand products instead of Western brand products. And, you know, this is part of the reason why. For example, you've seen this shift in the Chinese auto industry where it's gone from two thirds of all Chinese. All vehicles sold in China were Western brands. Even if they were built in China. They were brands like Volkswagen and General Motors. Yeah, we are, we are doing the same thing, you know, and it's flipped and now it's. I think that's almost 70%.
Leo Laporte [01:03:04]:
I think the less dependent we are on one another, the riskier our relationship.
Sam Abuelsamid [01:03:08]:
No, absolutely. Yeah.
Mike Elgan [01:03:09]:
That's a different question.
Leo Laporte [01:03:10]:
China, by the way, you know, we opened up Trump because Jensen Huang probably gave him a million bucks or whatever, said, okay, yeah, you go ahead, sell those Nvidia chips.
Sam Abuelsamid [01:03:20]:
But now, now they don't want them. But China's saying, because the Chinese government is telling Chinese AI Companies.
Leo Laporte [01:03:24]:
Do you keep your stupid chips.
Sam Abuelsamid [01:03:26]:
Yeah, don't use the Nvidia chips. Use, Use other alternatives.
Leo Laporte [01:03:29]:
Huawei is developing very, very good chips that they hope will compete. That's, to me, smart policy. I mean, look, there's a lot of things wrong about the prc, a lot of things wrong with, with, with the Chinese government. I'm just mostly interested in preserving relations with them as opposed to going to war with them.
Daniel Rubino [01:03:51]:
I agree with your. That the idea. They're defensive, too. China doesn't have a history of aggression like, say, Russia and the Soviet Union does. They're definitely interested in protecting their borders. They're definitely interested.
Leo Laporte [01:04:05]:
Well, to be fair, they have a history of aggression. World War II era.
Daniel Rubino [01:04:09]:
Right.
Leo Laporte [01:04:09]:
Where they pretty much conquered Southeast Asia.
Daniel Rubino [01:04:12]:
And, but now, now they, now they see themselves.
Sam Abuelsamid [01:04:15]:
China, at that point in World War II, prior to the war, had been occupied by the Japanese.
Leo Laporte [01:04:20]:
It was vice versa. Yeah. So the history of China has not always been isolation.
Daniel Rubino [01:04:25]:
No, no. A long time ago. But like, now they see themselves more as economic power, as a way with.
Leo Laporte [01:04:30]:
The belt and road initiatives. Yes.
Daniel Rubino [01:04:32]:
I don't really fear, you know, this idea of, like, China wanting to invade and Expand their empire kind of thing. They see it more as an economic.
Leo Laporte [01:04:41]:
Want to buy a BYD ev?
Daniel Rubino [01:04:43]:
I know, I do too.
Leo Laporte [01:04:45]:
Am I. You think that's silly, Sam?
Sam Abuelsamid [01:04:49]:
Not at all. I mean, they may make very good cars. Yeah, you can buy one.
Daniel Rubino [01:04:54]:
Aren't they about to cancel the Ford Lightning?
Sam Abuelsamid [01:04:57]:
Probably, yes.
Leo Laporte [01:04:58]:
What? Yeah, I know they suspended production.
Sam Abuelsamid [01:05:02]:
Yeah, they're probably, they're probably going to discontinue it because it has, it has not met up to the overly inflated sales expectations that they had a few years ago.
Leo Laporte [01:05:13]:
Well, they had the F150, the best selling vehicle in America, right?
Sam Abuelsamid [01:05:17]:
Well, no, actually the, the F series as a whole, including the heavy duty models, is the best selling overall. But the individual best selling nameplate is actually the Toyota RAV4. It outsells the F150.
Leo Laporte [01:05:31]:
So. But the F Series trucks are very successful. And so Ford probably thought maybe if we make an electric version of this, especially one that you could power your home with in a blackout.
Sam Abuelsamid [01:05:41]:
Well, and you know, they did.
Leo Laporte [01:05:42]:
It was very expensive. I think it was price, right?
Sam Abuelsamid [01:05:45]:
Yeah, well, there was a bunch of factors. You know, when they initially. Yeah, when they initially launched it, you know, they had it priced at a base price of $40,000. The problem is when they launched it in early 2022, you know, that was followed within a few months by the invasion of Ukraine, which then sent certain commodity prices like nickel and a few other metals skyrocketing, which caused battery prices to skyrocket as a result. And so they had to raise the prices on that. At the same time, you know, Ford had made some planning mistakes. They put in way too much. They, they grew the, the capacity way too much.
Sam Abuelsamid [01:06:25]:
You know, because when they originally were developing the Lightning, they were planning on selling about 25,000 a year. By the time it started production, they had increased that to 50,000 a year. And then a couple of months later, they raised that to 80. And then they doubled the size of the factory and raised it to 150. The problem is last year was their best sales year for the Lightning and they sold about 33,000 units.
Leo Laporte [01:06:46]:
Yeah.
Sam Abuelsamid [01:06:47]:
And so when, when you're running at about 20% of your capacity of your production capacity, you lose a lot of money.
Leo Laporte [01:06:55]:
Is that why really we're blocking Chinese car sales in the US Is protectionism?
Sam Abuelsamid [01:07:00]:
That, that is a big part of it is protectionism, because. Right. You know, the crush us. Yeah, yeah. Well, you know, the problem, the problem is, you know, the Chinese, by taking control of a lot of the supply chain for Key materials, you know, like lithium processing, graphite processing, a lot of the rare earth metal processing, you know, they've cornered a lot of that market. And so, you know, they control to a large degree the pricing on those materials. And so that gives them an advantage that we cannot compete with until we build up our own independent supply chain of those materials.
Mike Elgan [01:07:41]:
But here's a question for you. So it seems to me that if you balance different interests, you might consider the idea of just opening the gates to Chinese electric vehicles, letting them flood the US Market. If they're half the price for consumers, everybody's going to buy them. And what would that do to the quality of our air? And, and the, the, the number of, you know, the longevity of, of what.
Leo Laporte [01:08:07]:
Would it do to.
Sam Abuelsamid [01:08:08]:
It would absolutely, it would absolutely have environmental benefits for the United States. The problem is that we also have an administration that doesn't care. They don't, you know, they don't care. In fact, they are openly antagonistic to the, I don't know if any idea of improving.
Leo Laporte [01:08:24]:
At the cop, at the COP meetings, they didn't say anything about fossil fuel reduction.
Mike Elgan [01:08:29]:
But I'm not talking about what, nobody cares. What's the right thing to do? Is it to protect the industry or.
Leo Laporte [01:08:35]:
Is it to protect the auto? Should the American. Does the American automatic. You deserve to be protected, I guess.
Mike Elgan [01:08:40]:
Is the question, is it more important than health?
Leo Laporte [01:08:42]:
Right.
Sam Abuelsamid [01:08:44]:
You know, the, the, the thing is, you know, if, if we lose all that employment in the United States, all that manufacturing employment, who's going to be able to afford to buy those cars even if they're cheap? Who's going to, who's going to have the, the income to buy those vehicles even at half the price of the US Built models?
Leo Laporte [01:09:02]:
Which is why none of these car companies are looking at robotics production or reducing the number of workers on the line. Because they know that. No, that's not true.
Mike Elgan [01:09:12]:
I just wonder what percentage of a car, what percentage of a car comes from China anyway?
Sam Abuelsamid [01:09:19]:
Some, for the most part, it's actually not that big. It's probably less than about 10 or 15% in most cases.
Leo Laporte [01:09:29]:
And these are, these are, this is kind of a bastion of American industrial jobs, right?
Sam Abuelsamid [01:09:36]:
This is, this is, yeah, there's, there's about, there's about 3 million direct jobs in the auto industry in the United States. But the thing is those 3 million jobs also support about about seven or eight times that many. So about 20 to 25 million jobs in other, other businesses. You know, those, those people that work for automakers and suppliers, you know, down, down the value chain. You know, those people are spending money in other businesses.
Daniel Rubino [01:10:06]:
But so what you got to do then is you just open up BYD factories. Right. I mean, Honda did this, Toyota does. That's true. Yeah. I just find this whole discussion funny just because, you know, growing up I did political science and economics. It's just like I remember the free market and it was just what the chips land where they made you got to really compete. And then in the region, real world.
Leo Laporte [01:10:28]:
Apparently things get tight. Or a Kenzie Keynesian guy.
Sam Abuelsamid [01:10:32]:
Yeah.
Daniel Rubino [01:10:33]:
Well, I'm, I try to be a consistent person. So if I wanted to articulate, by the way, principle.
Leo Laporte [01:10:39]:
I think this is the single biggest problem we have in the United States is it's not consistent. Every four to eight years we completely flip.
Sam Abuelsamid [01:10:45]:
Yeah.
Leo Laporte [01:10:46]:
And do something different. So no one comfortable.
Daniel Rubino [01:10:48]:
We don't. It's politically unpopular and just like what happened to free market, man?
Leo Laporte [01:10:53]:
Like I thought that's what no one in our, in our, in our discord says. We've been trying to build a relationship with China since Nixon. They've taken the opportunity to pillage everything they can get their hands on. It's like trying to build a relationship with a crackhead who keeps robbing your house.
Daniel Rubino [01:11:07]:
They figured it out because there's the other thing too. Socioeconomics. You open up the markets and then they'll magically convert to democracy. Although this policy apparently could never apply to Cuba. Cuba was too radical and too big, you know, which I of China instead was just like, we'll keep the authoritarianism and have the capitalism.
Sam Abuelsamid [01:11:30]:
Yeah. They, they, they picked up the, they cherry picked the pieces that, that they wanted, you know, and they're making it work.
Daniel Rubino [01:11:35]:
I mean, yeah, look at, there are.
Leo Laporte [01:11:37]:
Problems with every single form of government. The real problem is, you said it earlier, humans.
Sam Abuelsamid [01:11:43]:
Yes. Yeah. We, we are the fundamental problem.
Leo Laporte [01:11:46]:
We're the problem.
Daniel Rubino [01:11:47]:
Yeah. So we got to get AI the, the thing.
Leo Laporte [01:11:50]:
Maybe that's why we, we need AI Somebody needs to run this.
Sam Abuelsamid [01:11:53]:
The thing about China, you know, the, the probably the best example of what has happened in China is that book Apple in China. You know, and, and that book talks about, yeah. Talks about how, you know, the Chinese circulated workers through the various Foxconn and, and Apple facilities in China and their supplier, their supply chain to learn how to build stuff and how to design stuff. And then those people went off and formed competitors like Xiaomi and Huawei and, and so many others. And you know, that's just in one industry. They've done that across the entire spectrum. They've done that in the auto industry. They've done it in every industry they compete.
Leo Laporte [01:12:30]:
Isn't that what the free market's supposed to celebrate is competition?
Sam Abuelsamid [01:12:35]:
Well, the problem, the problem, the fundamental problem with the free market argument, you know, if you go back to Adam Smith, you know, is, you know, you, there, there's, there's this idea of, you know, if you have enough buyers and sellers, you know, then you reach an equilibrium and you know, everybody has to be informed and you know, if you have a sufficient level of competition on both sides of that equation, you know, then you can, you can have a free market. But over the last 150 years, the global economy has shifted and it's grown, you know, the number of sellers has shrunk so dramatically in so many areas.
Leo Laporte [01:13:14]:
Right.
Sam Abuelsamid [01:13:14]:
That we, we don't really have the pieces in place that we need to actually have that kind of free market. So because, because we've allowed that level of consolidation, we have to, we have to manage it. You know, otherwise, you know, the, the small number of players on one side of that equation, the small number of sellers relative to the number of buyers limits the ability to actually have real competition.
Leo Laporte [01:13:41]:
Well, and there's a larger capitalism which is you basically end up creating an oligarchy because we have made it legal, which is what we've done super rich to.
Daniel Rubino [01:13:50]:
That's the irony of, that's the irony of capitalism. It, everybody talks about competition, but it's fundamentally hate competition.
Mike Elgan [01:13:58]:
Well, but, but also, I mean we, we've in the last.
Sam Abuelsamid [01:14:01]:
And the same is true on the other side.
Mike Elgan [01:14:03]:
But, but, but basically there's a difference between meddling in with the, with the market and taxing the rich. Like if we had a, if we had like a negative income tax proposed by some economics economists like 40, 50 years ago, where the more money your greater your net worth, the higher the percentage of your tax rate. If we did that sort of thing, okay, we, we would, we'd, we'd be in a much better position. And that's not meddling with the market, that's just redistributing wealth. Yeah, basically.
Sam Abuelsamid [01:14:32]:
And you know, when I said, you know, managing that, that economic equation, you know, that, that, that b. I wasn't necessarily advocating anything specific, but I agree with you, Mike, that you know, when, when you have too much of that influence on one side or the other, you know, then you need to start managing that somehow. And taxation is, I think, a very effective way of doing that.
Mike Elgan [01:14:56]:
Yeah, Capitalism is indispensable. What we have is a rotten political system. The most successful countries in the world with the best environmental records, with the most humane societies are capitalist countries. And some of the worst are capitalist countries as well. It's our politics we need to fix.
Leo Laporte [01:15:16]:
On that note, let's pause, let you absorb all of this great information. Were you all poli sci majors? Because it feels.
Mike Elgan [01:15:25]:
I was.
Sam Abuelsamid [01:15:25]:
I was an engineer.
Daniel Rubino [01:15:26]:
I was.
Leo Laporte [01:15:26]:
Okay, two out of three. And I was a Chinese major, so there.
Sam Abuelsamid [01:15:33]:
I took one economic economics class.
Leo Laporte [01:15:35]:
I always thought economics was fun, but the problem was that we study microeconomics. Or is it macroeconomic? We study economics, but it's in a test tube. And so it's not quite. It's much more complex and chaotic than what we study.
Daniel Rubino [01:15:49]:
It's a science.
Leo Laporte [01:15:51]:
Science. I did work in the Brookings Economics Library as one of my jobs in college. Talk about dismal. Very dismal library. I always thought that this is the dismal science. I get it now. We will have more with a great panel. Mike Elgin, Samobul Samid, great to have you.
Leo Laporte [01:16:10]:
And of course, Daniel Rubino from Windows Central. Our show today, brought to you by Zapier. I love Zapier. I use Zapier every single day. But the beauty of Zapier is it automates the work. So I don't even think about using Zapier. It just when, for instance, I mean, we do. There's all sorts of things I have the lights turn on when the sun goes down, things like that.
Leo Laporte [01:16:34]:
But I also use Zapier for my workflow. For instance, when I bookmark stories for covering on later on our shows, I bookmark it in a tool called Raindrop IO. Zapier watches for changes. I don't have to do anything. It sees that I've bookmarked, it automatically posts that story on our special social feed on Mastodon. At Twit Social, there's a Twitt News links. So post the new news link up there. It also automatically reformats the story and the headline and the URL in just the right way to put it into a Google sheet, which then the editors can copy and paste later to put into our show rundowns.
Leo Laporte [01:17:14]:
I mean, it's a lifesaver. I don't know how many millions of hours of my time has been saved over the last last five or six years, but Zapier has been a boon to me. Well, now Zapier is even better because Zapier has become an AI orchestration platform. You take the Workflows. You've already got the integrations with more than 3,000 tools that all the tools you use at work. But now you add access to the best AI models and suddenly you could bring the power of AI to any workflow. So for that workflow I just described, I could have, I could plug in Chat GPT and have it automatically take those stories, create a briefing book, a synopsis that we could send out. I don't do that.
Leo Laporte [01:18:00]:
I should really should do that. I mean, there's just all sorts of ways you can add AI to existing workflows or create new workflows entirely. You can connect the top models, they've got Claude, they've got Chat GPT and more. You can add AI wherever you need it, whether it's an AI powered workflow, an autonomous agent doing your work for you, a customer, chatbot, or anything else you can imagine. You can orchestrate it with Zapier. This is a transformational tool. I don't know about you, sometimes I look at these chatbots, I sit down and go, now what? You know, what do I do? It's too open. But with Zapier, I'm actually thinking about what do I need to get done? And then using Zapier plus AI to get it it done.
Leo Laporte [01:18:41]:
It is amazing. And you don't have to be a geek, you don't have to be a coder, you don't have to be a tech expert. Zapier is for everyone. Zapier should be a member of your team. Teams have Already automated over 300 million AI tests. 300 million using Zapier Join the millions of businesses transforming how they work with Zapier and AI. Get started for free by visiting zapier.com z-a p I-E-R.com twit zapier.com twit we thank them so much for their support. Casey Newton, who left mainstream media to create a, I think, very successful newsletter called Platformer.
Leo Laporte [01:19:29]:
He really made his name when Elon Musk bought Twitter a few years ago, but he still does some great stuff. This is a piece that he just published. Roblox is a problem, but it's a symptom of something worse. What is the role of tech journalism in a world where CEOs no longer feel shame? And by the way, it's not just CEOs, it's politicians, it's everybody in the public eye. Shame seems to be in shortage all over. He says he uses Roblox as the example. Roblox, of course, 151 million monthly users, most of them underage. A lot of people very concerned about the predators on Roblox.
Leo Laporte [01:20:13]:
And Roblox has had a kind of success of chain of failures with child safety, Casey says. On Friday, the Hard Fork team published our interview with Roblox CEO. In the days since, it's the most discussed interview we've done in three years. Journalists described the interview as bizarre, unhinged in a car crash. That was because Roblox CEO David Baszucki basically had a dismissive attitude towards child safety. Casey says. This unfortunately struck me as familiar that lately tech CEOs have basically just thrown up their hands and said, said, hey, this is just the way it is. How long are you guys going to be going on about all this? A lack of shame makes it hard for journalists, tech journalists in particular, to do our job.
Leo Laporte [01:21:17]:
You guys.
Daniel Rubino [01:21:17]:
I wonder what started that era, the lack of shame for doing.
Leo Laporte [01:21:21]:
You agree that that's. This is the era we're in though, right?
Daniel Rubino [01:21:23]:
Oh, 100%, yeah.
Leo Laporte [01:21:24]:
Mark Zuckerberg started it, I think, because.
Sam Abuelsamid [01:21:27]:
Part of it is because media turned these guys into rock stars.
Leo Laporte [01:21:32]:
Bingo. Yeah.
Sam Abuelsamid [01:21:33]:
You know, and so they, they, they, they, they, you know, we treated them like gods and so they're gonna act.
Leo Laporte [01:21:40]:
Like it over and over again. Casey writes, we've seen leaders in bouzouki's position choose growth over guardrails. Right. And when we journalists ask, why are so many of their users suffering? Executives.
Sam Abuelsamid [01:21:54]:
Laughter.
Leo Laporte [01:21:55]:
And tell us we're the crazy ones. Look at TikTok, look at Meta. And I really do think Mark Zuckerberg started this by, you know, break things, apologize later, you know, don't ask for permission, just do it. And then say, oh, I'm sorry, I didn't mean to do that. And we let him get away with it again and again and again and.
Sam Abuelsamid [01:22:16]:
Eventually they just stop, you know, saying, we didn't mean to do that.
Leo Laporte [01:22:19]:
The lesson learned was there's no, there's no consequence.
Mike Elgan [01:22:22]:
Well, there is a consequence and it's reward. So the incentives are all messed up.
Sam Abuelsamid [01:22:27]:
Yeah. No, negative.
Mike Elgan [01:22:28]:
We live in a world where, where politicians and business tycoons are rewarded for shameless grifting, for ignoring the needs of children. In this case, he's basically knowingly. And this is right out of the Meta playbook, exposing children to harm for the money. And he thinks that because there's so much money to be made, that's a reason why it's okay. That's the world we're living in in terms of our incentives, we need to change the incentives so that the incentive is to protect children. If you, to have a responsibility, if you have a children's platform, essentially. And we need to apply similar standards that we would apply to the real world. Just because it's digital and online doesn't mean that there are no rules.
Mike Elgan [01:23:15]:
Imagine if Chuck E. Cheese had a policy where they were encouraging adults to come in and have, like, private rooms where they could be alone with kids. What would we think? What would we do about that?
Sam Abuelsamid [01:23:27]:
Right.
Mike Elgan [01:23:28]:
We would do something about it. But if it's online. Well, it's okay because it's online. No, I think, I don't think it should be okay. And I think the incentives should be aligned with what the public wants, which is, you know, protection for kids.
Daniel Rubino [01:23:41]:
They're writing the COVID from the blowback with Twitter and this idea of like, regulation and, and people enforcing policies on platforms. And so while Roblox wasn't directly a part of that, that's the, like, mood of the, the atmosphere right now that we're in, which is that, oh, you don't touch companies, you let them do whatever they want, whatever the repercussions may be. And we don't put laws in into place. And so, you know, it's funny though, going back to the, the era of, you know, do whatever you want and not be apologetic. I. There's a good documentary on Netflix about Rob Ford up in Canada.
Mike Elgan [01:24:18]:
Yeah.
Leo Laporte [01:24:20]:
Cocaine issues. Didn't he.
Sam Abuelsamid [01:24:22]:
Yeah.
Daniel Rubino [01:24:22]:
He was, as far as I could tell, the first real person in, in pop culture media that did something wrong and didn't apologize for it. Just like, yeah, that's who I am, or video will come out. And he was like, that's not me.
Sam Abuelsamid [01:24:36]:
That's a lie.
Daniel Rubino [01:24:37]:
And all this. And it literally set the blueprint for Trump and everybody else and to come along and just be like, no, you're wrong. Prove it. That's not me. And all this, it's. But that's just a side thing I noticed. But that's the era we live in now.
Sam Abuelsamid [01:24:50]:
You know, brother is now the premier of Ontario.
Daniel Rubino [01:24:52]:
I know.
Leo Laporte [01:24:53]:
Isn't that ironic? Yeah. The interesting thing is, I mean, this is what the tobacco companies did for years, right? They said, oh, there's nothing wrong. They knew, knew they had all the evidence that cigarettes caused cancer. They denied it, they deflected it, they delayed it and they. But they didn't get away with it because the press and Congress persisted and finally brought them down.
Mike Elgan [01:25:17]:
And this is what the Democrats are hoping for, the midterms to Bring this issue to light. Because in fact, this divides the Republican Party. There are the, you know, pro Trump folks who are like, yeah, just let everybody, you know, let the business tycoons do what they want. But there are a whole bunch of Republicans who are like, no, no, we want strong, you know, protections and guardrails.
Leo Laporte [01:25:38]:
This is why you see those weird schisms, these dichotomies with Marjorie Taylor Greene and Steve Bannon who you think are like total MAGA supporters suddenly being, you know, apostate to the religion. It's like, what's going on?
Daniel Rubino [01:25:53]:
And all the people who are against the Epstein stuff and want those files.
Leo Laporte [01:25:56]:
Released was the beginning, wasn't it?
Daniel Rubino [01:25:58]:
Yeah, yeah, yeah. Those people should also be outrage, outraged about what's going on with Roblox. And these online companies loved Roblox favorite platform.
Sam Abuelsamid [01:26:09]:
Do we know that he didn't?
Leo Laporte [01:26:11]:
That's not true, is it? No, you're just. That's a joke, I hope.
Mike Elgan [01:26:14]:
Yes. A dark one, but, you know, dark joke.
Leo Laporte [01:26:17]:
Yes.
Mike Elgan [01:26:17]:
So, so Sarah Wynn Williams, who's a former senior policy advisor at Meta, wrote a book about, about Facebook and, and about their whole attitude and it's called Careless People.
Leo Laporte [01:26:30]:
Right.
Mike Elgan [01:26:30]:
And so, so we, we, we incentivize the careless people. I do disagree with sort of part of the tone of, of, of Casey's article. I mean, it's a great article. Everyone should read it. I really respect him and his, and, and so I'm not dissing him, but I do think that he is calling attention to by making it seem like it's very widespread. Actually, Meta and Roblox are among the worst, guiltiest parties in this world. I don't think they're typical. I don't think Meta is typical among, among social networking platforms.
Mike Elgan [01:27:07]:
I don't think Roblox is typical. There's way too much of it to be sure. And I think that's a big part of his point. But they're the worst. I mean, Meta is the worst. There's a. Has decades long history at Meta of them doing research, finding out that their product does harm, and then explicitly and deliberately choosing to do nothing about it because there's more money to be made and doing nothing.
Leo Laporte [01:27:32]:
Yeah. I mean, there's evidence galore that they knew and did nothing. This reminds me of the tobacco companies.
Mike Elgan [01:27:38]:
Yes.
Leo Laporte [01:27:39]:
Do you think, I mean, it took decades before the tobacco companies finally got kind of brought. And by the way, it's still completely legal to smoke cigarettes. So. So maybe they haven't been brought to their knees. Do you Think that there will be a comeuppance that at some point Meta and Roblox and companies like that will have to.
Mike Elgan [01:28:00]:
I do. And I think music. Yeah. I think the political actors who are defending the Robloxes and the Metas of the world are gonna get their asses handed to them. And I think that will trigger a congressional. Cuz that's what we need as Congress to actually show up and do their jobs.
Leo Laporte [01:28:19]:
Well, somebody else says we need a Federal Trade Commission that actually is empowered to do its job.
Daniel Rubino [01:28:24]:
And people who aren't financially trying to.
Sam Abuelsamid [01:28:27]:
Do their job, you know, they, they haven't been successful. I mean look at the, the ruling on Meta just a couple of weeks ago and you know, yeah, they're basically the courts.
Leo Laporte [01:28:37]:
Yeah, it's over.
Sam Abuelsamid [01:28:37]:
You know, they, you know, they've tried and just haven't been very effective.
Leo Laporte [01:28:44]:
Give.
Sam Abuelsamid [01:28:45]:
So you know, even when we had Lena Khan, you know, she didn't really achieve anything.
Leo Laporte [01:28:50]:
Yeah, that was her action against Meta, which the court just said, yeah, you didn't prove your case, so we're dismissing it.
Mike Elgan [01:28:58]:
You have to remember and look at the big picture and remember that the current political era is a personally, you know, it's a one person sort of movement and when that one person exits the stage, the movement is essentially going to at the very least change very radically. I think within a decade we'll be back to a somewhat more normal politics where, where we'll be able to regulate these things, where states will have more rights and, and you can see, you know, a combination of Europe and California.
Leo Laporte [01:29:35]:
I hope you're right.
Sam Abuelsamid [01:29:36]:
Yeah, unfortunately I'm not as confident confident as you, Mike, you know, because you know, you look at what, you know, okay, what happens after this guy's gone, you know, then junior, you've got the, you know, the tech oligarchs, you know, that are, that's to me really trolling. It's not about everybody behind him, it's.
Leo Laporte [01:29:55]:
About the power we've reached. I wonder if we reached a critical mass where these companies are so big, so rich, so powerful that nobody can stop them.
Mike Elgan [01:30:04]:
Well, a big part of their power, particular leader right now is being given to them by the administration. And the administration has the power that it has because of the number of voters who are personally loyal to that individual person. When that, when that is broken, then we're, it's a different ball game. Yeah, they have way too much power and so on, but right now that's being put on steroids by the administration.
Leo Laporte [01:30:33]:
Which is problematic. Because it's going to be a point where it doesn't matter.
Sam Abuelsamid [01:30:39]:
By the time he's gone, you can't fight. We have so much, you know, this going back to the whole AI thing, you know, you've got so much slop being generated by AI that you know, even now, I mean, nobody knows what's true or false anymore, right?
Mike Elgan [01:30:55]:
But the, the thing that I always like to go back to in my writing of opinion columns is that we always forget that the more extreme things get in any direction, the more the extreme will be the reaction. And so there's a lot of disgust around AI slop. For example, it's somewhere between 70 and 90% by the, by the end of this year, right next month and within 3, 4, 5, 6 years, people, some are predicting 99 of the content online will be AI generated. Okay, that's not just this one way street, this one way trend. There'll be a whole body reaction to this that won't be positive. It's already happening, of course. And little one by one companies are starting to build in tools to be able to turn off AI, to be able to reject it. This is the problem with flooding the world.
Mike Elgan [01:31:53]:
Like look at the podcast industry. There are podcasts companies so quote unquote, podcast thousands of 5,000 shows, right? It's just garbage, right? AI with AI. Exactly. They just, it's on autopilot. It takes, cost them $3 per episode. Can you imagine, Leo?
Leo Laporte [01:32:09]:
But they don't have to make much to pay for that.
Mike Elgan [01:32:12]:
Exactly, exactly.
Leo Laporte [01:32:14]:
And you're done.
Mike Elgan [01:32:15]:
Right, but, but what, what is the, what is the right policy for a podcast app? The right policy is, is give the people what they want and let them turn this garbage off.
Leo Laporte [01:32:25]:
Look what's happening to Spotify. Lisa, who was it, was it Lisa who had a Uber driver that said, yeah, this is just a sideline. What I've been doing lately is using AI to create. I do an album a week of AI music which I post and I make right now a few hundred dollars. But if one of them hits, I won't have to drive anymore. He is single handedly publishing hundreds of albums a year of AI music on Spotify. And I think that there's probably thousands.
Sam Abuelsamid [01:32:56]:
Of people like, well even, even Spotify is probably is.
Leo Laporte [01:33:00]:
Spotify's doing it themselves.
Sam Abuelsamid [01:33:01]:
That's right, much of it. So they don't have to pay anybody for it. They can prioritize that stuff in playlists that they generate so they don't have to pay anybody else.
Leo Laporte [01:33:11]:
It's Good. Benito's on vacation because I think he'd be pretty upset right now as a musician. He hates, hates Suno. I think Suno is pretty amazing. Amazing. They just by the way made a deal with Warner Music for AI Music.
Daniel Rubino [01:33:29]:
The good news is, sorry, just like, you know what everybody's saying here with Mike especially there, I think could be definitely a pushback against this. And this is the one nice thing about open and free markets is a company can come along, create another music or podcast platform, specifically ban anti stuff. Yeah, yeah. And that's their marketing, right? This is pure music, this is original and it would attract a market. Just like how vinyl made a huge comeback, right. Because people got sick of the whole digital experience and they're like wanting more authentic. And vinyl still does really well, especially for collectors. So there's opportunities here to sort of really push back, back against this.
Daniel Rubino [01:34:12]:
Same with the social network stuff, which finally seems to be. I believe this is one of the first years where Twitter and Facebook and all that are starting to slowly drop in numbers. But there could be opportunities here to create another platform that is what people want and where they want to be versus what has been forced down their throats. And everybody thinks it is because all this stuff was great years ago. Now it's just like horrible.
Leo Laporte [01:34:35]:
My kid says that he knows a lot of gamers who say we'll never play a game with AI in it, right? But Tim, but Tim Sweeney says, yeah, go ahead. Because guess what? Every game will have AI in it. Pretty soon they'll all be AI. It doesn't matter what you say today, it's happening.
Sam Abuelsamid [01:34:55]:
I mean every game for I. I don't know how long has had non player characters.
Leo Laporte [01:35:00]:
Right.
Sam Abuelsamid [01:35:00]:
It may not be.
Leo Laporte [01:35:01]:
Well, imagine how good they'll be.
Sam Abuelsamid [01:35:03]:
They won't gener. It may not have been created by, you know, current types of generative AI, but it was, it was done procedurally.
Leo Laporte [01:35:10]:
How many times did you hear that guy say, yeah, I was, I used to be a fighter, but then I got an arrow through the knee over and over and over again. Now there'll be unique, interesting stor vibrant stories coming from your npc.
Daniel Rubino [01:35:22]:
This comes down to transparency. I think gamers will be okay with some amount of AI, but the companies just need to be forthright about how they're using. Using it.
Leo Laporte [01:35:32]:
Hollywood was very. Was up in arms over what was the. The name of the movie about the architect that they used AI to fix their Hungarian accents. The mi. The most minor use of AI ever. And as a result, even Though it was nominated for an Academy Award, it didn't win. I think it didn't win because Hollywood said, no, no, we got to punish these guys because it was.
Daniel Rubino [01:35:54]:
I think that stuff will be normal though. Stuff like that I think is fine. But it's going to do get a lot. These problems are going to get a lot more difficult. Like I said, when AI writes the first Academy Award winning script and it can make its own movies and all this kind of stuff and people can't tell a difference then the real moral quandaries.
Leo Laporte [01:36:12]:
Well, what are the biggest. I think we're getting very close to that by the way. I think, yeah, we're inches away from that.
Mike Elgan [01:36:17]:
What are the biggest movies in the last three. Three years, right, is Maverick, you know Top Gun. Maverick.
Leo Laporte [01:36:22]:
Yeah.
Mike Elgan [01:36:23]:
And F1. And the reason everybody loves them them is they put, they, they lashed actors into real jets and real cars and they hit the gas and everybody loved it because it was real interesting.
Daniel Rubino [01:36:36]:
There's a lot of good YouTube analysis right now talking about cinematography, how much it's changed in the last 10 years. There's a video going on right now about. They call it the Netflix lighting about how there's no shadows anymore in the way they shoot stuff. And it, and it has this fakeness to it that doesn't. It's fascinating stuff because in this, you know, desire to make mass media and make as much as possible, it's also become very generic. And so yeah, Mike's right. Like people want to see. You see it all the time with special effects discussions where people miss even movies from the 80s where they use practical effects.
Daniel Rubino [01:37:12]:
There's something that people enjoy a lot.
Leo Laporte [01:37:15]:
I think we're always going to be able to tell the difference that we'll like be able to distinguish.
Daniel Rubino [01:37:20]:
It does become that.
Mike Elgan [01:37:23]:
Honestly, you hit the nail on the head. There are two arguments. One is librarians in Michigan are banning all AI books in libraries because the AI books tend to be garbage. What happens when they're wonderful and fantastic? So that's another issue. But the second thing is that will it matter to people whether they're being communicated with and sharing the. The experiences of another human being? Or will it not matter? Would they just want to be plugged in the Matrix and just receive stuff coming from the machine? I tend to think that once we're super, when we, once we stop being dazzled by the incredible advancements that are being made every day in AI and, and this happens very quickly, by the way. Remember when Nano Banana came out, it was like Three days where everybody's like, wow, this is like everything's different. And now we're just kind of like, yeah, whatever.
Mike Elgan [01:38:11]:
Once we're stopping dazzled by these things, we're going to kind of want something real. We're going to want something that comes from another person. We're going to want to, we're going to remember that content is basically writing and photography and movie making and all these things which exist for people to share their experiences.
Sam Abuelsamid [01:38:30]:
I'd like to think that you're right, Mike, but you know, then when you look at, you know, how many of the, you know, best selling books, you know, best selling novels, you know, are basically dreck, you know, already. Yeah.
Leo Laporte [01:38:45]:
Danielle Steel could be. It probably be better if human talking.
Sam Abuelsamid [01:38:49]:
About, you know, the, the what, what's the vampire series? Twilight. You know Twilight? Yeah. And you know, Danielle Steel and you know, so many of the, you know, you know, Daniel, the Da Vinci Code, you know, stuff, you know, a lot of this, you know, these are all written by human beings.
Mike Elgan [01:39:09]:
Yeah.
Sam Abuelsamid [01:39:09]:
It's not exactly, you know, great literature.
Daniel Rubino [01:39:13]:
Of course people are going to read in the future.
Sam Abuelsamid [01:39:15]:
Yeah. You know, people are, people are willing to absorb a lot of not great stuff anyway. So I don't know that, you know. Yeah, I mean, I, I, I definitely would prefer that, you know, to have, you know, real experience, you know, to share in real human experience. But I think most people don't really.
Mike Elgan [01:39:37]:
It may be a luxury for the rich and the well educated. I mean, so many things are like that. For example, we, we replaced horses with cars and replaced, you know, sailboats with the rich stuff. But actually if you're rich, you have a horse and a sailboat. Right. So, and a fireplace. We have central heating now, but you.
Daniel Rubino [01:39:52]:
Know, that's what's happening with college. That college is going to be just for rich people as almost as a luxury to get that extra education to be worldly. While for normal people there's not going to be any financial benefit in doing that.
Mike Elgan [01:40:08]:
So ladies and gentlemen, if you like good literature, make sure you're rich.
Daniel Rubino [01:40:11]:
I don't know. I managed to turn a political science degree into something else.
Leo Laporte [01:40:16]:
It was your connections, man. Where'd you go to school?
Daniel Rubino [01:40:19]:
Sienna.
Leo Laporte [01:40:21]:
No, that wasn't your connections then.
Sam Abuelsamid [01:40:22]:
Okay.
Mike Elgan [01:40:23]:
No, my sister went to Worcester College, so.
Leo Laporte [01:40:26]:
Worcester, Worcester College.
Mike Elgan [01:40:28]:
California girl.
Leo Laporte [01:40:29]:
Yeah. I was just down the road from PC and Uri. I had a friend who always was very proud of his degree from Uri, the university.
Daniel Rubino [01:40:37]:
PC was my second, my second school. I was going to my Backup PC.
Leo Laporte [01:40:41]:
Good school, good basketball power. Let's take a break. We're talking with Siena graduate Daniel. Daniel Rubino. I'm sorry, Sienna. We work with the New York Times. I'm not knocking Sienna.
Daniel Rubino [01:40:52]:
I'm not knocking Sienna. New York Times Times poll. That's the one. Most famous polls out.
Leo Laporte [01:40:56]:
That's right.
Daniel Rubino [01:40:56]:
That's what we used to do.
Leo Laporte [01:40:57]:
It's the same Sienna. That's Sienna.
Daniel Rubino [01:40:59]:
Yeah. Sienna. Yeah.
Leo Laporte [01:41:00]:
There you go.
Daniel Rubino [01:41:01]:
It's a very famous political.
Leo Laporte [01:41:03]:
Yeah, well, it's the place for poli sci. Okay, now we're talking. Sam A bull. Samet, where'd you get your engineering degree?
Sam Abuelsamid [01:41:12]:
When I was there, it was known as the GMI Engineering and Management Institute. Now it's known as Kettering University. Kettering in Flint.
Leo Laporte [01:41:18]:
Okay, cool. Now another well known name. Sloan Kettering. Yeah. And Mike Elgin.
Mike Elgan [01:41:25]:
Ucla. Ucla.
Leo Laporte [01:41:26]:
Ucla, baby.
Mike Elgan [01:41:27]:
From. From. For a Twitter audience, I think the most important thing about UCLA was it's the place where the world's first email was received. And in fact, I remember the room we'd walk by, it's like, oh, that's the room where the first email. They typed one word and the whole thing crashed.
Leo Laporte [01:41:41]:
And that's where the at sign was invented.
Mike Elgan [01:41:43]:
Exactly, yeah.
Leo Laporte [01:41:45]:
Actually Postel was at usc, so nevermind. Oh, yeah.
Mike Elgan [01:41:47]:
This bastard.
Leo Laporte [01:41:49]:
Ucla, baby. We're gonna take a break, come back with more, lots more to talk about with our great panel. But our show today, brought to you by Melissa, one of our favorite sponsors. Many years, many, many years we've been telling you about the trusted data quality expert. They've been doing it longer than we have though, since 1985. And they've got lots of plaudits and kudos. Recognized by G2, for instance, as leaders in both, both Data quality suite and global address verification. Melissa's trusted by businesses worldwide to eliminate costly errors.
Leo Laporte [01:42:23]:
You know, I mean, you. You send something to the wrong address, that is money down the tubes to boost efficiency, even to drive growth. Pipeline safety, one of those things that often goes unnoticed until something goes wrong. Right. The Paradigm alliance exists to prevent incidents by keeping. This is a really important use of address information. Keeping communities informed and pipeline operators compliant. It's very important in their business.
Leo Laporte [01:42:53]:
They need to find and reach the right people fast. You don't want a wrong address or phone number in your database. With Melissa, Paradigm got an extension of its team with a faster access and more intelligent decisions. President of the Paradigm alliance said, says quote, Melissa's team doesn't just provide data. This is what I was saying. They provide solutions tailored to our unique challenges. While address verification, of course is the foundation of Melissa's services, Melissa's Data Enrichment Services do a whole lot more. Organizations can use MELISSA to build a more comprehensive, more accurate view of their business processes by using MELISSA as part of their data management strategy.
Leo Laporte [01:43:32]:
Give you an example, the Parkinson's Institute and Clinical Center. They've amassed an invaluable collection of clinical data over the last 30 years. It's unmatched worldwide. The relationship between MELISSA and PICC has opened up many transformative use cases including improved understanding of how genes, proteins and treatments impact Parkinson's disease. Thanks to Melissa, Pick can do earlier identification of candidates for clinical trials or alternative treatments. Very important, important to their research and by the way, PICC is able to do secure and by the way, very lucrative data sharing with collaborators with the help of Melissa. PICC's CEO says quote, Melissa has made it possible for us to transform our complex and diverse data into a unified research ready knowledge resource, end quote. See, that's what I mean.
Leo Laporte [01:44:26]:
Melissa is they're data scientists, they're experts in data of all kinds. And of course your data is completely secure with MELISSA compliant, safe with melissa solutions. They're GDPR compliant, they're CCPA compliant, ISO 27001 certified. They meet SOC2 and HIPAA high trust standards for information security management. Get started today with 1000 records cleaned for free@melissa.com TWIT that's Melissa. M E L I S S A We thank them so much for their support of this week in tech. Back to the show we go. Let me go over to our rundown here.
Leo Laporte [01:45:14]:
Oh, I did want to talk about this bleeping computer. Had kind of a scoop looking at the code in OpenAI's ChatGPT Android bot. Actually this is from an X post by somebody named Tibor. There are now references to an ad feature in Chat GPT, including something called Bazaar B A Z A A R, content wrapper, search ads and a search ad carousel. Now, you know we've seen this kind of stuff in source code. Doesn't mean necessarily they're going to roll it out, but it kind of implies that maybe ChatGPT is looking at putting ads into their results. What do you think? I mean this is an inevitability though, given the costs of doing this stuff.
Mike Elgan [01:46:03]:
It's gonna happen for sure and they know more about you than Google does and it's going to be floated with ads. Yeah, right.
Leo Laporte [01:46:13]:
We survive. I mean, I don't use Google for that. Reason I use Kagi, but, but like I pay, I pay the price for that. That's a luxury. I'm a privileged, you know, person, you.
Sam Abuelsamid [01:46:23]:
Know, the, the cost of running these data centers, running, you know, running these AI systems is just too high. You're not, you're, you're not going to be able to get enough people to pay the way you do Leo, for, for all of these services.
Leo Laporte [01:46:37]:
Of course, I mean, 25 bucks a month for ad free searches.
Sam Abuelsamid [01:46:40]:
Yeah. You're not, you're not going to get enough customers for all of this stuff. You're going to have to generate revenues elsewhere. Just as we've seen in media streaming and TV streaming. You know, they, you know, once they started doing ad based services at a lower cost, you know, now they generate more revenue, they, or they have higher profit margins, they have higher margins from their ad based services than they do from the customers that are paying for the ad free services.
Mike Elgan [01:47:08]:
I don't, I don't think it's that simple though because just because they have to pay back their investors and, and pay for their expensive data processing and so on doesn't mean that we have to pay them for the, we can go elsewhere. We're talking about Chinese open source models that are far cheaper. They're going to be models that are free. I think they're going to have a hard time raising the money they need to raise. It's not automatic. They're going to have to really work at it and be very, very clever to come up with monetization schemes that don't, you know, sort of turn everybody away.
Sam Abuelsamid [01:47:41]:
But even those cheaper Chinese models, you know, they've got to generate some revenue somewhere and they can't afford to just give away open source models forever. At some point they're going to have to get some revenues to pay. Even at their lower costs, they're still going to have to pay off those costs at some point.
Mike Elgan [01:48:01]:
I also wonder to what extent the public will continue to go open their browser, open a web app to use chatbots when it comes to if, if smart glasses become much more ubiquitous, if watches that give you information, this kind of information, other wearables, if that might sort of divert how we use these models and if the cost of using them might be built in. I mean, I, I really don't know. If you're using like Ray Ban metaglasses and you figure you're trying to figure out how, when you're just, just talking to a voice, how you monetize that with advertising, that's That's a, that's a tricky.
Daniel Rubino [01:48:43]:
Yeah.
Leo Laporte [01:48:44]:
You don't want it to be stuck in the results. Right.
Mike Elgan [01:48:46]:
Yeah.
Leo Laporte [01:48:47]:
So you know, if I'm looking for sneakers, you don't want the results to be tainted by an ad like it.
Sam Abuelsamid [01:48:53]:
Is at Amazon or Google or anywhere else.
Leo Laporte [01:48:57]:
Never mind. Forget.
Daniel Rubino [01:48:58]:
I.
Leo Laporte [01:49:00]:
Don'T mind ads as long as they're clearly posted as ads. Right?
Mike Elgan [01:49:04]:
Yeah. Yeah. That's just so much money to be raised to pay for this stuff. It's just.
Leo Laporte [01:49:10]:
Well, one of the things OpenAI is spending money on, $3 billion is this partnership with Johnny I've to make some thing.
Sam Abuelsamid [01:49:22]:
They're bringing back the, the buttonless ipod shuffle.
Leo Laporte [01:49:25]:
Yeah. Who knows, right?
Sam Abuelsamid [01:49:27]:
Rounded with rounded edges.
Leo Laporte [01:49:28]:
Now, Johnny Ivan, Sam Altman on stage with Steve Jobs widow actually. Lorraine Powell Job Jobs at the Emerson Collective demo day talked a little bit about this device. They say, oh, it's, it's no more than two years off. They didn't really say what the device is but, but Sam Altman said, I'm sorry Johnny, I've said I love solutions that teeter on appearing almost naive in their simplicity. I also love incredibly intelligent, sophisticated products that you want to, to touch and you feel no intimidation and you want to use almost carelessly, that you use them almost without thought that they're just tools. Sam Altman said, he wants it so good you could eat it. You want to eat it.
Sam Abuelsamid [01:50:16]:
The buttonless ipod shuffle, I'm telling you.
Mike Elgan [01:50:19]:
Yeah, exactly. I mean the thing that concerns me here, I mean Johnny, I've is happy to. To take this $.5 billion company IO oh wow. And you know, yes, I would take that as well. But the problem is that at Apple, where of course Johnny Eve did his best work, he had this sensibility. The kinds of things he's saying about this secret product are the kinds of things he would say about Apple. But at Apple there were all these engineers and visionaries and Steve Jobs himself saying okay, we're going to do this iPhone and it's going to do this and it's going to do that. And, and Johnny, I've came in to sort of do the sort and fit and finish type stuff and, and make some.
Leo Laporte [01:51:01]:
But he had to make a handle on them built by the engineers that Steve Jobs wouldn't go, that's crap.
Mike Elgan [01:51:07]:
Right. But, but, but now he's to OpenAI which has no experience in building consumer electronics. They're saying hey Johnny, I've you're the hard way guy builds the next iPhone and he's. He's not. That's not.
Sam Abuelsamid [01:51:18]:
Well. And we, we saw how that played out when he didn't have that editor and, and Steve Jobs at Apple, you know, you ended up butterfly keyboards and all kinds of other nonsense.
Mike Elgan [01:51:29]:
I also believe there's a, there's a delusion that you can do wearables. AI Wearables that are not watches or glasses. And I don't believe it. Nobody's gonna be something. Nobody's gonna lash something around their neck that those just. The human factor is never gonna happen.
Daniel Rubino [01:51:44]:
It's just another, we got enough stuff to charge already. Also, also, you know this point about like, no one to go and say, I want to make the next iPhone. Like anytime you go into a problem saying that's what we're going to do, we're going to create the next big thing that's like you're already attacking it from the wrong way. You should be asking, like, you know, what problem are we trying to solve with this? Or what opportunity is there that we can leverage? And right now, I mean, listen, maybe they'll create something that'll just totally blow our minds. I'm open to the idea, but yeah, the idea of like a pendant or something else, I got to put on something else. I got to charge and then. And things with microphones, it's going to be listening to me.
Sam Abuelsamid [01:52:30]:
Yeah.
Daniel Rubino [01:52:31]:
Maybe it doesn't have a display.
Sam Abuelsamid [01:52:32]:
Definitely don't want to be talking to everything, you know.
Daniel Rubino [01:52:36]:
Right.
Sam Abuelsamid [01:52:36]:
Using a voice interface for every interaction.
Mike Elgan [01:52:39]:
Right. If they want to knock it out of the park. The designer types like Jony. I've should be working on fashion. So what we need are glasses that have a little display that you. That can look at hand gestures and things like that. So it's not all.
Leo Laporte [01:52:53]:
Is that the meta spray band display? I mean, that's pretty close to what it.
Mike Elgan [01:52:56]:
Pretty close to it. But the problem is nobody's gonna. Nobody, nobody except geeks like me would wear that thing. Right. So, so what they. What they should be pouring all their energy behind is creating a line of, of a. Of 200 different styles of glasses. This is a huge opportunity for designers, Right.
Mike Elgan [01:53:15]:
Because now they're into high fashion at high prices with high tech. Right? That's the opportunity. It's gonna be glasses. Don't try to invent something else. It's going to be glasses.
Daniel Rubino [01:53:28]:
And that's why Apple attacked it from the wrong angle with the Division Pro. It was like an Apple.
Leo Laporte [01:53:34]:
Even Apple says, you know, maybe we'll do glasses.
Mike Elgan [01:53:37]:
They're coming out with glasses. Eventually, but they'll be really late to market market with, with open AI. Just give me open AI and glasses and they, they would, they would immediately have a successful product if the glasses were wearable.
Leo Laporte [01:53:49]:
You turned me on, Mike, to the meta Ray Bans. You've been a fan of those.
Mike Elgan [01:53:52]:
Yeah. If they can do better than, than Ray Ban with glasses, that, that's a win.
Leo Laporte [01:53:57]:
Yeah. The problem with these is it's meta and I don't. And I feel like I'm giving.
Daniel Rubino [01:54:01]:
Exactly.
Sam Abuelsamid [01:54:02]:
It's like.
Leo Laporte [01:54:04]:
Right.
Daniel Rubino [01:54:05]:
No one wants meta on this.
Sam Abuelsamid [01:54:07]:
That's why I won't buy them.
Leo Laporte [01:54:08]:
Right.
Daniel Rubino [01:54:09]:
Yeah.
Leo Laporte [01:54:09]:
But is open AI better or. I mean how, how do we know they're not going to be exactly the same?
Daniel Rubino [01:54:14]:
We just need competition though, right? Every company.
Leo Laporte [01:54:19]:
I would, I mean maybe that's foolish, but of all the companies that could do this, Apple seems the least likely to, to misuse my information.
Mike Elgan [01:54:30]:
Right, but. And they're also in a good position. They're very design centric, they're very, you know, fashionable.
Leo Laporte [01:54:36]:
They know how to make hardware, they know how to make products.
Mike Elgan [01:54:39]:
Look at what they did with watch bands. Watch bands are the most boring thing you can imagine and, but they went crazy with it. You can have this one, you can have that one. Here's a $2,000 one. They should do that with glasses and I think they'd have a hit on their hands. Anything that is AI that you can wear all day and look good doing it is going to be a successful.
Leo Laporte [01:55:01]:
And you think the interface will be voice?
Mike Elgan [01:55:03]:
I think the main interface will be voice, but also hand gestures. Some of the things are doing Vision Pro. Right. Can be eventually built into glasses that can be worn all day. But some of the, some of the voice input can be offloaded to a visual screen and some of the voice output can be, you know, hand gestures.
Leo Laporte [01:55:24]:
Even with the medical. Now you could say what am I looking at? And the meta AI will do its best. And in many cases it's pretty good at describing what you're looking at or translating exactly.
Mike Elgan [01:55:34]:
I think we're very close. But if you sort of like go down the dark alley of some wearable that isn't glasses or a watch, then it's just a wasted opportunity.
Leo Laporte [01:55:44]:
I was wearing, as many of you know, I was wearing for a long time, a variety of different pendants that were recording everything that would happen and would give me AI synopses of the day and to do lists and stuff. And I found it actually somewhat useful. But I also found that people were A little skittish when they found out I was wearing that. I think they'd even be more skittish if they saw me wearing glasses with a tally light blinking as I was recording the scenario or talking to or talking back to me. People are a little concerned about privacy.
Mike Elgan [01:56:15]:
Yep. I think there's a huge opportunity for privacy. Privacy based AI. So AI glasses are going to be huge. People will be seriously annoyed by the cameras and the recording. And so. So whoever comes out with a really compelling, good looking pair of glasses that give you the AI interaction without the camera people is going to, you know.
Daniel Rubino [01:56:36]:
Always the opportunities here. Opportunist. There's the anti AI type market where. And there are now glasses. Right. That you wear that fool. Or supposed to be able to block facial recognition systems. Right, Right.
Daniel Rubino [01:56:49]:
So there is that alternate. You know, you can hack it from another problem. But there's a lot of potential innovation there too that I think people are going to be interested in because. Yeah, you know, guarantee your own privacy. Right.
Leo Laporte [01:57:01]:
Here's a mask that blocks AI face recognition from every angle. Sure. Just walk around with that on. Actually, Neal Stephenson in one of his novels, Fall or Dodge in Hell, actually people are walking around town with these masks on. And one of the things these masks can do is have different Personas. So you can choose the Persona that you want to display to the AIs and cameras all around you. So you know, today I'm going to be the mild mannered tech reporter, tomorrow I'm going to be a punk rocker. And you get to choose what it.
Daniel Rubino [01:57:37]:
I mean that mask, like, you know, if you're making like a sci fi movie in the future, that was kind of techno dystopian.
Leo Laporte [01:57:44]:
Yeah.
Daniel Rubino [01:57:44]:
People would be wearing those. So yeah, market.
Leo Laporte [01:57:48]:
Right, Right.
Daniel Rubino [01:57:49]:
So like I don't know, let's make that future happen. Like the cyber truck. Like I know people like dump on it. But like again like if you were thinking, I'll give, you know, Elon Musk credit if you were thinking of movie like 30 years in the future, what a pickup would look like. Sure. Right. That's a prop for a movie. So he just made it, that's all.
Daniel Rubino [01:58:06]:
But, but I don't know, I think people should do these things.
Leo Laporte [01:58:08]:
Article in the AP today. One tech tip. Modern cars are spying on you. Here's what you could do about it. Sam, you've been telling me this for years. Is, are all modern cars spying on.
Sam Abuelsamid [01:58:19]:
Us to some degree or another? Pretty much, yeah.
Leo Laporte [01:58:23]:
This is from Kelvin Chan, who's almost.
Sam Abuelsamid [01:58:25]:
All new Cars now have built in data modems. Yeah.
Leo Laporte [01:58:30]:
You know, but that's a feature for US that's like OnStar or.
Sam Abuelsamid [01:58:33]:
Well, it's a feature, you know, it's a feature that we can benefit from, but it's also a feature that the manufacturers benefit from. You know, I mean GM, you know, started this in 1996 when they launched the original version of OnStar and it's gotten significant. Progressively more sophisticated over.
Leo Laporte [01:58:49]:
They got in trouble. They were banned for five years from disclosing data collected from drivers to consumer reporting agencies. They had been doing that.
Sam Abuelsamid [01:58:57]:
Yeah. Well other manufacturers are doing it too.
Leo Laporte [01:59:00]:
Too.
Sam Abuelsamid [01:59:00]:
They're, they're sharing information with companies like Verisk and Octa that are, you know, that they, they collect, they aggregate information about driver behavior.
Leo Laporte [01:59:12]:
Yeah. Whether you're driving insurance, modeling, driving at night.
Sam Abuelsamid [01:59:15]:
And they're sharing it, you know, directly with insurance companies.
Daniel Rubino [01:59:18]:
That's the real like. Yeah. I think people, you know, when it comes to advertising, people just accept it. Right. It's like, all right, they're going to sell my data to some advertiser. Whatever, whatever. But this idea of like it's going to insurance companies and all of a sudden people's insurance rates are going up because of the way they're driving. And but, but that is, isn't that.
Leo Laporte [01:59:38]:
How it should be? If you're a crappy driver, why should I be paying. You should pay more for your insurance.
Mike Elgan [01:59:44]:
Yeah, but that's not, that's not something that I'm going to opt into. Right. As a consumer.
Leo Laporte [01:59:48]:
That's why they don't have this.
Mike Elgan [01:59:50]:
My car, my car is supposed to be on my side. This is not.
Sam Abuelsamid [01:59:54]:
Insurance pools are supposed to work is you, you aggregate the risk across the entire population.
Leo Laporte [02:00:00]:
But insurance companies do all they can to figure out which kind of like if you're a teenage boy you're going to pay more because you're not. Because you are a bad driver. But just because statistically you're more likely to be progressive.
Daniel Rubino [02:00:11]:
Has an opt in system. Right. Where you actually use a sensor.
Sam Abuelsamid [02:00:15]:
Yeah, yeah. And so every insurance company will either give you a dongle or on most newer cars you can, you can opt into sharing your data directly with your insurer instead of using it and they.
Daniel Rubino [02:00:28]:
Give you at least the, the care of will lower your, you're stand a chance of a lower premium. So like at least.
Sam Abuelsamid [02:00:34]:
Or that there's also the opposite side of that too. Your insurance premium could go up, which is what happens with Tesla. If you get Tesla insurance, they will actually Adjust your premiums not just every six months or every year, year, but from month to month. So if in any given month you accelerate too hard too many times or hit the brakes too hard too many times, the next month your insurance premiums can go up significantly.
Daniel Rubino [02:00:59]:
And that almost makes sense too because that's what they also use to get people into the full self driving betas, right? Is if you're a good driver, you get early access to those features. So but I've also heard good things from Tesla insurance in terms of, of the pricing. But.
Sam Abuelsamid [02:01:16]:
Well again, it depends if, if you are a very careful driver, you can save money on it because you know, with other insurance companies, Teslas are expensive to insure because they are very expensive to repair. And so the, the insurance, the insurance premiums for those vehicles tend to be higher. And if you are a, a very careful driver, you can save money with Tesla insurance. That's why they created it. But then they also realized they had so much data that they could also make a lot more money by raising premiums significantly on drivers who had even slightly higher risk factors in their driving behavior.
Leo Laporte [02:01:53]:
So this AP report has a link to a vehicle privacy report from Privacy for Cars. So I just entered my VIN and it tells me what my car is collecting. It's not collecting biometrics, but it is collecting a user profile, location identifiers. It sells it to, doesn't sell it to insurance agencies, but it does sell it to the government. Oh, that's good. And then it, which nice. They give you some action buttons so you can file a privacy request. You can ask government officials to act.
Leo Laporte [02:02:29]:
I think that's kind of. Is this, do you know if this, Sam, is this something you're familiar with? Is this, is this legit?
Sam Abuelsamid [02:02:35]:
Yeah, I mean, I don't know how, it's hard to say, you know, how effective it really is, how much you can really protect your privacy with tools like this. Right. Really the only way that you can truly protect your privacy is by, you know, going in and turning off the data modem entirely.
Leo Laporte [02:02:53]:
But then you may not want to do. What are the downsides to doing that?
Sam Abuelsamid [02:02:56]:
Well, you, you lose access, potentially lose access to a lot of services. So a lot of vehicles today, you know, increasingly manufacturers are opting to, opting into using Android Automotive as the, the operating system for their infotainment systems. And they've got Google Automotive Services. So you got Google Maps, you got Google Assistant, you got the Play Store so you can download apps right in there. So you don't have to connect your phone to the vehicle. You know, which is, you know, the smart smartphone projection like CarPlay and Android Auto is very problematic. You know, oftentimes, you know, you will, your phone will get disconnected while you're driving. And you know, it' not, it's not very, it's not a very reliable system oftentimes.
Sam Abuelsamid [02:03:39]:
And so having it built into the vehicle is a much more robust system. But again, it doesn't work unless your vehicle is connected to the network.
Daniel Rubino [02:03:47]:
Right.
Sam Abuelsamid [02:03:47]:
And also, you know, manufacturers, when you buy one of those vehicles, they will give you some period of time when you get complimentary connectivity. Usually it's anywhere from one to three years. Although GM on their EVS, where they don't even have projection, CarPlay or Android Auto available, they actually give you eight years of connectivity. That means that for the first eight years your car is going to be connected to GM's network. And they can also harvest data, you know, and some of that is, is, you know, it's not just data that they share to third parties, but some of it is also data that they share about how customers are using their vehicles. And they feed that into their product development process. Figure out, okay, what features are people actually using as opposed to, you know, I mean, we put these things in there, but are they actually using them? And you know, so, so then, you know, maybe in future models they, they eliminate the stuff that nobody's actually using. So there's, there's upsides and downsides to it.
Sam Abuelsamid [02:04:41]:
But you know, really the, the only way, you know, is to just disconnect the vehicle entirely or just, you know, drive an old car like I do.
Leo Laporte [02:04:48]:
Yeah, it does. They do have some opt out features, right? It depends the state you live in. I imagine in California there is quite a bit of opt out.
Sam Abuelsamid [02:04:56]:
So it's, it's mostly all in or all out. Oh, you, it's, it's usually not feature by feature.
Daniel Rubino [02:05:04]:
Okay.
Sam Abuelsamid [02:05:04]:
It's, you know, you can, you can have the service or you turn it off entirely.
Mike Elgan [02:05:09]:
Do you think that's a kind of dark pattern that.
Leo Laporte [02:05:11]:
Yeah.
Mike Elgan [02:05:12]:
Tries to.
Leo Laporte [02:05:12]:
Because you want the benefits like I want the maps to work. Yeah, right.
Sam Abuelsamid [02:05:16]:
Yeah. I mean, you don't. Without it, you know, you won't have any navigation.
Daniel Rubino [02:05:19]:
Right.
Sam Abuelsamid [02:05:20]:
You won't have any.
Leo Laporte [02:05:22]:
I don't know how to drive anymore. If I didn't have the gps, I wouldn't, I wouldn't get anyone somewhere. I'm just like, now what, where do I turn? What? You, you sent me a, a fun video of the original Ev.
Sam Abuelsamid [02:05:37]:
Yeah, before, before we get to that, I want to say, you know, one other thing. Last week I was talking to a company that is doing some interesting stuff. So a lot of new cars, you know, and Leo, you first saw this when we drove that Cadillac CT6.
Leo Laporte [02:05:50]:
Yes.
Sam Abuelsamid [02:05:50]:
The Super Cruise. All those years it was watching driver. More and more new cars have driver monitor systems.
Leo Laporte [02:05:56]:
Yeah.
Sam Abuelsamid [02:05:57]:
That are, you know, looking to see, you know, when you're using systems like this, they want to see, you know, are you actually, you know, they want to make sure you're actually watching the road and keep, you know, paying attention so you can take over control. And some cases, even some cars that don't have hands free capability are including that now, including those driver monitor cameras to look for driver distraction or drowsiness, you know, so they can give you alerts, you know, that, hey, okay, maybe, you know, if you're, if it looks.
Leo Laporte [02:06:23]:
Like you're my yells at me if I pick my nose, it says you're not paying attention. Yeah, now that seems a little intrusive.
Sam Abuelsamid [02:06:33]:
Well, you know, I mean that. If you can't pick your nose in.
Leo Laporte [02:06:36]:
Your own little damn car, what, where can you.
Sam Abuelsamid [02:06:38]:
Yeah, I mean, you know, if you're, if you're driving while you're tired, you know, that's as dangerous as being impaired.
Leo Laporte [02:06:44]:
I agree.
Sam Abuelsamid [02:06:45]:
And so, you know, having something that can alert you to that and say, hey, you know, maybe it's time to pull over and take a break, you know. But this company is utilizing that capability along with the connectivity and AI and, and also GPS location of the vehicle so that, you know, what you can do now is, you know, if you see something, you know, if you see a business or you see a sign, you know, and you say, you know, and I forget what the keyword is, but you say a keyword and say what is that? That, and then it'll grab your current location and it uses that camera to figure out your gaze, to figure out what you're looking at from your current location. And it will give you information about what that is, what that business is. And so traditionally these driver monitor systems have been completely localized. They weren't sending that data anywhere. It was just being used in real time to monitor the driver state. But now they're starting to use that to figure out, okay, what are you looking at at? And I can definitely see where, you know, it starts to go, you know, beyond that, even without you asking, it can be, it's for you figuring out what you're looking at, what's, what's getting your attention as you're driving.
Mike Elgan [02:07:57]:
And that's something cars, Cars are getting really like, sort of Orwellian. In the last. I don't know when I started noticing this, maybe in the last year, whenever I rent a car in Europe, the. If you're 1km over the speed limit, it will beep at you.
Sam Abuelsamid [02:08:14]:
Well, that's, that's, that's a new European regulation.
Mike Elgan [02:08:17]:
Yes. And it's a nightmare.
Sam Abuelsamid [02:08:18]:
Intelligence speed is 1, what, 1 mile.
Leo Laporte [02:08:21]:
Over the speed drives.
Mike Elgan [02:08:22]:
If you're as much as 1 mile an hour over the speed limit, it will beep at you until you go below it again.
Sam Abuelsamid [02:08:28]:
Usually they do that in Singapore.
Leo Laporte [02:08:30]:
I remember my Singaporean taxi driver going, don't worry about it, all the way down.
Mike Elgan [02:08:37]:
It's so annoying.
Leo Laporte [02:08:38]:
But, but I don't think anybody in America drives the speed limit.
Mike Elgan [02:08:44]:
No.
Leo Laporte [02:08:45]:
People?
Sam Abuelsamid [02:08:45]:
Well, it depends, you know, I mean, if you're, if you're in Manhattan, you know, the average speed's only eight miles an hour.
Leo Laporte [02:08:49]:
Oh, yes.
Sam Abuelsamid [02:08:50]:
If you can't drive if you're on the 405.
Mike Elgan [02:08:52]:
Yeah.
Leo Laporte [02:08:52]:
I'm just saying when, when people can.
Mike Elgan [02:08:55]:
Yeah, they.
Leo Laporte [02:08:56]:
I mean, this is one of the.
Mike Elgan [02:08:57]:
Best reasons to live in the United States is you're not harassed by. In Australia, they have a law that says if people in the back seat don't have their seatbelt, seatbelt, the seat belt thing beeps. You know, that's probably a good idea. But still, you add up all these, all these different things and the car is just, just nagging you constantly about the lane. You get too close to the lane, there's something near you, you're over the speed limit. It's like, what, are you married to it?
Sam Abuelsamid [02:09:22]:
I was driving something last week, you know, and I was. All I wanted to do is just reposition the car in my driveway. And, you know, sometimes, you know, if I'm repositioning, I mean, normally I always wear a seatbelt, but sometimes if I'm just moving a, you know, I, I will hop in, you know, just put it in reverse, back it up, move, and just move it over without necessarily buckling up. And it would not even go into driver reverse until I put my seat belt on.
Mike Elgan [02:09:49]:
Yeah.
Sam Abuelsamid [02:09:49]:
It would not leave park.
Leo Laporte [02:09:50]:
My, my car does that. It doesn't, it doesn't put the seat in position or anything until I fasten my seat belt. And then everything got us. Okay, now we can drive.
Daniel Rubino [02:10:00]:
Yeah. I mean, it's like I, I agree. You know, this is all. Has potential for abuse and I think that's this is always a downside with all this stuff. It's like on one hand it's extremely helpful. Automotive deaths are huge in the United States. They're, you know, drunk driving, just poor driving. Like a lot of that is just common.
Daniel Rubino [02:10:21]:
And so this is a way to combat that and save a lot of lives. It's also the potential to save money on insurance and repairs and everything else. The problem is, is that all we never see the savings ourselves because everything is for profit. So like whatever savings insurance companies are going to get from this back to the shareholders and we're not going to get it. Because right now if you go online and despite all this technology and everything that's happening, all you see are people complaining about car insurance rates going up by a lot. A lot of people's, their rates are just going up and up all the, across the country, no matter how old they are. And it's just like no one is getting the benefits from any of this in terms of lower costs. And that's where really like really bothers me is that we're not seeing the benefits you get.
Daniel Rubino [02:11:08]:
Make the same arguments with like health, right. So we're all wearing smart bands and measuring our sleep. And so like in theory all that information could go to your doctor and they could lower your insurance because you're a healthier person, but they won't, you know, and so then it's just like, you know, that's where I don't know.
Mike Elgan [02:11:28]:
I see it as a part of the transition to self driving where, where AI is, is in control of the vehicle. Because when you're actually driving around one of these cars that's beeping at you because of the speed limit, it, it's superseding your judgment. So if I'm on a, if I'm on a quiet country road in the middle of the night where I know there's no traffic around, I can see for 10 miles ahead of me and I want to go 1km over the speed limit, it won't let me because that's machine thinking. Right. So I'm, I'm actually functioning as an AI just following this algorithm.
Leo Laporte [02:12:02]:
That's right.
Mike Elgan [02:12:03]:
Instead of human judgment.
Leo Laporte [02:12:04]:
Yeah.
Mike Elgan [02:12:05]:
And so, and so one by one, these little features will chip away at our individual autonomy to the point where at some point we give up all autonomy. We just get in and we, we. The, the, the car drives us.
Sam Abuelsamid [02:12:19]:
My 36 year old car doesn't have any of that.
Mike Elgan [02:12:22]:
Beautiful.
Leo Laporte [02:12:23]:
Drives a Miata.
Mike Elgan [02:12:25]:
Oh, I used to drive Miata.
Sam Abuelsamid [02:12:26]:
They're wonderful first generation.
Mike Elgan [02:12:29]:
Oh, nice.
Daniel Rubino [02:12:30]:
Until an SUV rolls over it.
Sam Abuelsamid [02:12:32]:
Well, yeah, but you know, at that, at that point I won't care because I'll, I'll be dead anyway. I'm, I'm well aware of the risks of that's the right animal driving that car. You know, I know I am almost entirely unprotected in that thing.
Leo Laporte [02:12:45]:
You're watching this week in tech, Daniel Robino, Sam Apples, Samit who lives life on the edge, Mike Elgin, and yes, me. So glad you're here. Especially our Club Twitt members. Thank you for supporting this show. We really appreciate it. Our spot. We'll have more in a bit, but first a word from our sponsor, Zscaler, the world's largest cloud security platform. The potential rewards of AI.
Leo Laporte [02:13:09]:
AI. I mean, we know about those. They're too great to ignore. But there are also risks if you're using AI in your business. Loss of sensitive data. There's also the issue of attacks against enterprise managed AI. Generative AI increases the opportunity for threat actors too, helping them rapidly create phishing lures. Get this, there were 1.3 million instances last year of Social Security numbers leaked through AI applications.
Leo Laporte [02:13:37]:
Not intentionally, just exfiltrated. In effect, ChatGPT and Microsoft Copilot saw nearly 3.2 million data violations in the same time. But you want to use AI in the business and you want to defend against AI driven attacks. Well, that means it's time for a modern approach with Zscaler's Zero Trust plus AI. It does all that. It removes your attack surface so you're less vulnerable, it secures your data everywhere, it safeguards your use of public and private AI and it protects against ransomware and AI powered phishing attacks. Check out what Shiva, who is the director of security and infrastructure at Zuora, says. We've got a great video about using Zscaler.
Leo Laporte [02:14:19]:
Watch.
Sam Abuelsamid [02:14:20]:
AI provides tremendous opportunities, but it also.
Leo Laporte [02:14:24]:
Brings tremendous security concerns when it comes to data privacy and data security.
Mike Elgan [02:14:28]:
The benefit of Zscaler with ZIA rolled.
Sam Abuelsamid [02:14:30]:
Out for us right now is giving us the insights of how our employees.
Mike Elgan [02:14:34]:
Are using various gen AI tools. So ability to monitor the activity, make sure that what we consider confidential and sensitive information according to, you know, companies.
Sam Abuelsamid [02:14:44]:
Data classification does not get fed into.
Leo Laporte [02:14:47]:
The public LLM models, et cetera. With Zero Trust plus AI, you can thrive in the AI era. You could stay ahead of the competition and you can remain resilient. Resilient even as threats and risks evolve. Learn more@zscaler.com Security that's Zscaler.com Security we thank him so much for the support of this week in tech. I just wanted to mention, I started talking about the EV one, which was the first ev. In fact, wasn't there a great documentary about how General Motors killed the ev?
Sam Abuelsamid [02:15:23]:
Who killed the electric car?
Leo Laporte [02:15:24]:
Killed the electric car?
Sam Abuelsamid [02:15:25]:
Car.
Leo Laporte [02:15:26]:
Somebody has. There is one privately owned GM EV1 out there and there's a, there's a YouTube video all about it.
Sam Abuelsamid [02:15:37]:
Yeah. So when, when GM launched the EV1 in 1997, they did it in response to the first round of California Air Resources Board mandates for.
Leo Laporte [02:15:47]:
Seems like a good idea.
Sam Abuelsamid [02:15:48]:
EVs to be sold. Yeah, yeah. You know they, in like 1989, they had, had debuted a concept called the Impact and it got so much.
Leo Laporte [02:15:59]:
Okay, first of all, terrible name for a car. I just want to say impact. Really? Crash 2000. I mean, come on.
Sam Abuelsamid [02:16:07]:
Okay. Yeah. It got such so much positive attention that they decided to go ahead and put it in production. And so when they finally launched it in 1997, you know, what happened was they, they, they never actually sold them to consumers. They only leased them.
Leo Laporte [02:16:27]:
Ah.
Sam Abuelsamid [02:16:28]:
And they, they built them for about five years and they built about 1100 or a little more than 1100. Tiny number. Yeah. And you know, they were, you know, the original ones, the first generation ones used lead acid batteries. But.
Leo Laporte [02:16:45]:
Well, there's your problem right there.
Sam Abuelsamid [02:16:47]:
Well, but that's, that's what was available.
Leo Laporte [02:16:48]:
Available. How much capacity could you have?
Sam Abuelsamid [02:16:51]:
They had a range of about, well, a rated range of 78 miles by the current standards. The way they test them now, it equates to about 55 miles. Yeah. So they could go depending on how you drove it. It was, it was about, you know, 50 to 70 miles.
Leo Laporte [02:17:08]:
Do they charge faster than lithium ion?
Sam Abuelsamid [02:17:10]:
No, no, but, but it was, it was technology that was available at the time, you know, and then they switched to nickel metal hydride, which is, is what Toyota was using in their hybrids. And you know, lithium ion batteries just weren't, it wasn't practical at that time to put into a vehicle.
Leo Laporte [02:17:27]:
Right.
Sam Abuelsamid [02:17:27]:
It wasn't until several years later that the first lithium ion battery EVs were produced. So they, they only leased these things and they, you know, they were, these things were, you know, I mean it was, it was an incredibly sophisticated car for its time, but they were basically hand built and most of the parts in this thing were completely bespoke. They used very few part were shared with other GM vehicles. And you know, they, they were running into a problem where, you know, they were building these things by hand, essentially, at. At a factory in Lansing, Michigan, that ironically is called the GM Craft Center. And so they. In 1993 or 94, I think they decided to withdraw them because they. They decided not want to continue to support these things because there were too many potential things that could go wrong with these cars.
Sam Abuelsamid [02:18:22]:
You know, they were. They, like I said, they were really sophisticated. The owner, the. The. The drivers loved them. But there were a lot of challenges with it, with supporting these things because there were so many bespoke components, and so much of it. So much of the car was. Was, you know, was, you know, kind of.
Sam Abuelsamid [02:18:40]:
Each vehicle was to. Was, to some degree, totally unique from all of. So, you know, they went to the. The. The customers that were leasing these things and said, okay, you got to give them back to us now. And they sent most of them to the crusher to avoid them being. Being kept out in the wild and being used, you know, with. Without proper support.
Sam Abuelsamid [02:19:07]:
They kept about 40 of them, and they, quote, unquote, decommissioned them, which means they took out the batteries and they did something, you know, they cut some. Some cables and so on and did some other things to make sure they couldn't actually run. And then they. They gave these to various academic institutions, universities, you know, to use for research and study. And one of these, you know, some of them are, you know, a lot of them are in museums now. One of these went to Clark Atlanta University, and it was used for a while. Obviously wasn't driven, but engineering students studied this thing. And then it basically got set aside and forgotten about.
Sam Abuelsamid [02:19:54]:
And so it was sitting in this parking lot, the storage lot at Clark Atlanta University. It had. It had been in a garage that had been. The garage was damaged in a storm. Part of the roof caved in. It shattered the windshield. They moved this thing outside. It was sitting in a lot.
Sam Abuelsamid [02:20:13]:
It had been sitting there for several years, essentially abandoned. And somebody from the security department of the school said, what's the deal with this thing? Nobody knew what was going on with it. It wasn't being used for anything. And so they called in a towing company and said, here, tow this thing away. And so the towing company got it. And under the laws in Georgia, you know, if, you know, they. They post, you know, they. They post online that, hey, you know, this car's here.
Sam Abuelsamid [02:20:43]:
If it's yours, come and clean, claim it, and, you know, pay the. Pay the storage fee and you can take it away. Well, nobody did. And, you know, the, the loan agreement when GM gave it Gave these things to the schools was that they could not sell them or give them away to anybody. You know, and GM could recall them at any time. They never did. But they, they, they had the ability, the legal ability to do so. But under Georgia law, because this storage company, the vehicle been declared abandoned, the.
Sam Abuelsamid [02:21:12]:
The towing company company towed it away. It was. They did all the things they're supposed to do with posting it, you know, saying, hey, you know, come and claim it if it's yours. Nobody did. So they put it up for auction. And so it was on this auction site and it ended up selling for $104,000. Yeah, it was.
Daniel Rubino [02:21:31]:
What.
Sam Abuelsamid [02:21:31]:
It was not drivable. Had no, no battery in it.
Leo Laporte [02:21:35]:
And so it made for perfect. You know, they probably made more than that on their YouTube channel, right?
Sam Abuelsamid [02:21:39]:
Probably. Yeah.
Leo Laporte [02:21:40]:
And.
Sam Abuelsamid [02:21:41]:
And so, you know, there's now been a couple of videos about this thing. I spent a couple of hours over the last couple of days watching the latest video.
Leo Laporte [02:21:48]:
This is the Questionable Garage, which is a great name for a car channel.
Sam Abuelsamid [02:21:53]:
Yeah. This guy that does this channel, the Questionable Garage, he's done all kinds of weird restorations on various cars.
Leo Laporte [02:21:59]:
He's.
Sam Abuelsamid [02:21:59]:
He did a GM S10 EV that was also built in the late 90s. They built a like about 150 of those. And he's got one that's actually. He's got running.
Leo Laporte [02:22:10]:
Hey, Mike, maybe when your son KE Kevin gets tired of his hydrogen vehicle, he could sell it to these guys for others.
Mike Elgan [02:22:16]:
Great idea. The future looks grim for hydrogen. But he still drives it.
Leo Laporte [02:22:20]:
He does.
Mike Elgan [02:22:21]:
He's driving it at this very moment from Los Angeles.
Leo Laporte [02:22:24]:
He's got a. He's got a Toyota. Yeah. Mirai.
Sam Abuelsamid [02:22:29]:
You got a first gen or a second gen Mirai?
Mike Elgan [02:22:31]:
I don't know.
Sam Abuelsamid [02:22:32]:
Is it the ugly one or the.
Mike Elgan [02:22:33]:
No, it's beautiful.
Sam Abuelsamid [02:22:34]:
Okay.
Mike Elgan [02:22:35]:
That's the generation.
Leo Laporte [02:22:36]:
And. And because he lives around the corner from a hydrogen filter filling station.
Mike Elgan [02:22:40]:
Yeah. Here in the Bay Area. They're everywhere. And it makes sense. Down the street. Yeah.
Leo Laporte [02:22:46]:
Didn't you tell me he gets free hydrogen for life?
Mike Elgan [02:22:49]:
No, not for life. It was for a bunch of years as part of the incentive to get him to buy it. Yeah, it was like six years or something like that.
Sam Abuelsamid [02:22:55]:
So this video, they're starting the restoration. They're basically taking the car completely apart to clean everything, fix everything.
Leo Laporte [02:23:03]:
Are they going to get.
Sam Abuelsamid [02:23:04]:
The goal is to eventually get it running. Yes.
Daniel Rubino [02:23:06]:
Wow.
Sam Abuelsamid [02:23:07]:
Wow. And you know, one again, because there's so many unique parts on this thing, like that windshield, they can't just call Safelight and say, hey, come and replace the windshield on this EV1. Because there's no stock of these things.
Daniel Rubino [02:23:19]:
Right.
Sam Abuelsamid [02:23:20]:
So what they, what they did was they, they, they got some matte vinyl wrap and put it across the windshield and put on a bunch of dots and they got a laser scanner and they made a 3D model of it. So they're going to try and get somebody to fabricate a windshield.
Leo Laporte [02:23:36]:
And this is what YouTube's great for because, because this is just some shade tree mechanics probably, but they can monetize it. They can actually make this happen. You don't have to go to NBC and say, give me $8 million for. We'll make a TV show about this thing. And they'd never get the money. This is fantastic.
Sam Abuelsamid [02:23:55]:
Yeah. But as you watch this and you see taking parts off, you see things like there's parts where they, you know, where there was obviously an interference clearance, like by the front doors, you know, there was a bracket that you can see where it was actually cut by hand with tin snips. So that, so it didn't interfere with the door when they opened the door. And there's other places where you can see where they whack it with a hammer to get the roof panel to fit.
Leo Laporte [02:24:20]:
It's handmade.
Sam Abuelsamid [02:24:21]:
Yeah.
Daniel Rubino [02:24:23]:
They have the opportunity here to do the funniest thing ever, which is just put a ice engine in it and just make it a regular car. Insult everyone. Yeah.
Sam Abuelsamid [02:24:36]:
You know, I mean, this is the only EV1 in private hands anywhere.
Leo Laporte [02:24:41]:
Wow.
Sam Abuelsamid [02:24:41]:
And so they, they want to get.
Leo Laporte [02:24:42]:
It back via illegal loophole. They can. GM doesn't get it back.
Sam Abuelsamid [02:24:46]:
Yeah, exactly.
Daniel Rubino [02:24:47]:
Yeah.
Leo Laporte [02:24:48]:
Wow.
Sam Abuelsamid [02:24:48]:
So it's pretty cool.
Mike Elgan [02:24:49]:
There's.
Leo Laporte [02:24:50]:
It.
Sam Abuelsamid [02:24:50]:
I mean, if you have any interest in evs, especially early evs, it's, it's worth, you know, like I said, it's about. This one was about two hours. But it's, it's fascinating to watch and see some of the things that GM does, did on this car. You know, creating, I mean, it's all, you know, the structure.
Leo Laporte [02:25:07]:
Do you think in. In all honesty that their intent was to create a market for an electric vehicle?
Sam Abuelsamid [02:25:14]:
Yeah, I think so.
Leo Laporte [02:25:15]:
They thought there was something there.
Sam Abuelsamid [02:25:17]:
Yeah. You know, I mean, they spent a lot of money and spent a lot of time developing this vehicle. I mean, it is, like I said, it is a really sophisticated vehicle, especially for the 1990s.
Leo Laporte [02:25:29]:
The time wasn't right. That's all.
Sam Abuelsamid [02:25:30]:
Yeah, it was. The time wasn't right. The technology wasn't Right. Yet they didn't have the battery technology.
Leo Laporte [02:25:35]:
A lot like these E glasses, these AI glasses, we're just not quite there yet. We just.
Sam Abuelsamid [02:25:42]:
But you know, by 2004, 2005, you know, now, you know, AC propulsion.
Leo Laporte [02:25:47]:
I'm saving these because maybe I'll have a YouTube channel. I can resurrect the meta.
Sam Abuelsamid [02:25:51]:
Yeah, There was a California company called AC Propulsion that worked with GM on the, the propulsion system for the EV1. And you know, they were, they were doing conversions of other cars to EVs and they were the first ones to actually do, do a lithium ion ev, you know, and they built this little sports car from a kit car. They took a kit car and modified it and put. Built a lithium ion battery pack for it. And that car was called the T0. And a couple of guys named Martin Eberhard and Mark Tarponing, you know, and found this thing and they got a test drive in it said, oh, we, we want, we want the, this we, you know, build kind of what Elon.
Leo Laporte [02:26:34]:
Did with the Lotus Elise. Right.
Sam Abuelsamid [02:26:35]:
Well, except Elon didn't actually do that.
Leo Laporte [02:26:38]:
Somebody. The original founders.
Sam Abuelsamid [02:26:41]:
Martin. Yeah, Martin. That's Martin. Martin Everhard and Mark Tarponing.
Leo Laporte [02:26:44]:
Oh, you're talking about them.
Sam Abuelsamid [02:26:46]:
Yes. And I'm sorry, I didn't mean to.
Leo Laporte [02:26:49]:
Cut to the punchline.
Sam Abuelsamid [02:26:50]:
Didn't want to build.
Leo Laporte [02:26:51]:
Apologies.
Sam Abuelsamid [02:26:52]:
So they said, okay, we'll do it.
Leo Laporte [02:26:53]:
Sounded familiar.
Sam Abuelsamid [02:26:54]:
They licensed some technology from those guys and they went to Lotus and got a chance chassis and had that modified and that eventually became the first generation Tesla Roadster.
Daniel Rubino [02:27:06]:
I'm sorry.
Sam Abuelsamid [02:27:08]:
With funding. You know, he was, he provided funding.
Mike Elgan [02:27:11]:
Right.
Sam Abuelsamid [02:27:12]:
But you know, it was Mark who.
Leo Laporte [02:27:13]:
Were the founders, but he also, to his credit, he saw that and said, you know, there could be a business there.
Sam Abuelsamid [02:27:19]:
Yeah. Oh no, absolutely. I mean, you know, if, if he had not put in that $30 million initial investment.
Daniel Rubino [02:27:26]:
Right.
Sam Abuelsamid [02:27:26]:
The Roadster never would have, never would have gotten off the ground.
Leo Laporte [02:27:29]:
I remember when we were back in the cottage 15, 20 years ago, one of our listeners had a roadster drove. It came out Dick D. Bartola and I got in, went for a ride. It was tiny little thing. That Lotus Elise is not a big car, but that was pretty exciting.
Sam Abuelsamid [02:27:46]:
Little did I know. Well, actually the, the Elise is smaller than the roadster. The Elise is so small that, okay, they actually stretched, they extended the Elise chassis by about 8 inches.
Leo Laporte [02:27:57]:
Remember, the seats are a little tight.
Sam Abuelsamid [02:27:59]:
Yeah, they were.
Leo Laporte [02:28:01]:
I wanted to ask you, Daniel, if we should be worried about RAM prices. You know, we talked about the fact that AI is driving power consumption, but it's also thanks to GPUs driving RAM prices through the roof. Some RAM kits, this from Ars Technica Andrew Cunningham, are over three times as expensive as they were 90 days days ago. Is this something we should worry about?
Daniel Rubino [02:28:26]:
Yep, it's. It's also expected to impact obviously the laptop market for next year. Yeah, all OEMs I'm hearing are expecting prices to go up.
Leo Laporte [02:28:38]:
Lenovo stockpiled a bunch of chips to avoid this, but will that.
Daniel Rubino [02:28:43]:
So Lenovo. Lenovo is the largest PC maker on the planet, so that makes. Makes sense. And they're smart for doing that. You know, they get a lot of their business also to through enterprise, so they have a lot of customers that rely on them for that stuff besides just the consumer market, so they can do that. There's already reports that Microsoft hasn't done this for Xbox, which is why there seems to be a dearth of sales for Xbox consoles this year, despite Sony offering some sales because Sony looks like they paid attention and planned ahead. On the other hand, it's ironic that Microsoft is doing cloud gaming, which is using all those ram chips and GPUs, thereby driving up the prices so much that they can't afford them now for the Xbox consoles. Yeah, I think this is going to be an interesting problem.
Daniel Rubino [02:29:40]:
I don't know expectations for when prices could come down. But yeah, that's why if people are interested in getting a laptop or need one, you may want to actually take a serious look at some of the sales going on. Because right now some of the current stock is really good. And you know, I've seen the new Qualcomm chips, don't get me wrong, they're really good and they're making some huge strides. As someone who runs a Surface Laptop 7 Surface Pro 11 with Snapdragon chips, I don't feel like I need, need a new processor at this point, which is kind of a first.
Leo Laporte [02:30:14]:
They did announce it at the. In Hawaii. They're going to this next generation.
Daniel Rubino [02:30:18]:
Yes.
Leo Laporte [02:30:19]:
Elite. The X2.
Daniel Rubino [02:30:20]:
Yeah. And they're pretty massive jumps in performance and for some people they're going to really benefit. But this is sort of the Apple thing. A lot of people who bought the.
Leo Laporte [02:30:27]:
M1, do they not use RAM like Apple does? Because Apple.
Daniel Rubino [02:30:31]:
So the new X X2 does, does.
Leo Laporte [02:30:34]:
Does it still get hit by the cost of DDR5 going up?
Daniel Rubino [02:30:38]:
Right. So, you know, I'm hearing the, some of the new laptops, I believe, have 48 megabytes of RAM. Is like the minimum gigabyte. These X2s or gigabytes because of. Because of the. Shared with the GPU and everything else. So.
Leo Laporte [02:30:54]:
So they're not buying NAND chips or are they? I mean, what's, what is. Is still a NAND chip in there, isn't it?
Daniel Rubino [02:30:59]:
Yeah, it's still nand, just solid still.
Leo Laporte [02:31:01]:
So it's still going to affect the price of even unified memory. Even Apple.
Daniel Rubino [02:31:05]:
Yeah.
Mike Elgan [02:31:06]:
Yeah.
Daniel Rubino [02:31:06]:
So I expect all those to go up in price.
Leo Laporte [02:31:09]:
There's a. We. We always. I used. In fact I did a video for them back in the day on how to build your own computer. There's a great. One of the last remaining PC component stores in San Francisco down on Market street called. Or actually it's on the.
Leo Laporte [02:31:21]:
It's on Mission or Harrison called Central Computer. And it's gotten so exp. Gotten so expensive now they don't have a price on the ram. They do mine market pricing for the Verge.
Sam Abuelsamid [02:31:33]:
They sell it like lobster price for the fish. Yeah.
Leo Laporte [02:31:36]:
You know when you go to a restaurant and the menu says market price, when you order, they're selling it at. They don't put price tags on there because of the price increase. We can't display fixed prices on certain products at this time. Just ask. You have to ask. That's wild. That's wild. So this is good to know.
Leo Laporte [02:31:57]:
See, I've been. I'm kind of holding off until next year's Apple MacBook Pro. That's going to be an M6 and it's going to have an OLED screen. It. It's. It seems like this. I have. I'm pretty satisfied with the M3 I've got.
Leo Laporte [02:32:09]:
So I'm thinking if I wait till next year. But am I going to. Are we going to see big price increases?
Daniel Rubino [02:32:14]:
I think there's definitely a possibility for it. Yeah. But now how much is big? Right. If it's one or two hundred dollars extra, that's okay.
Leo Laporte [02:32:21]:
I can live with that.
Daniel Rubino [02:32:22]:
For Apple people, I feel like that's just, you know, they're going to charge a lot more.
Leo Laporte [02:32:27]:
Yeah.
Daniel Rubino [02:32:28]:
Right, right. But for, you know, for laptops that are like the 6, $700 range. Right. This impacts things. And Apple is rumored to be bringing a much more affordable laptop to the market.
Leo Laporte [02:32:38]:
Right.
Daniel Rubino [02:32:39]:
Which I think this would.
Leo Laporte [02:32:40]:
And the rumor is going to have intel make the chip.
Daniel Rubino [02:32:43]:
Right. Yeah.
Leo Laporte [02:32:44]:
Which is wild. It'll still be a. It won't be an intel chip, it'll be an Apple. Apple designs Apple Silicon.
Daniel Rubino [02:32:49]:
Yeah. Intel Core 3 coming out they're doing better. Don't count intel out. I, I realize yeah, they have a lot of potential. This idea between x86 and arm and that x86 is doomed, is definitely overhyped. You know, X, there's nothing inherently in design of an X86 that makes it completely inefficient to ARM.
Sam Abuelsamid [02:33:13]:
But if I'm going to buy an.
Leo Laporte [02:33:15]:
X86, I'm going to buy an AMD, not an Intel. Right. Is that foolish of me?
Daniel Rubino [02:33:20]:
I, I, it depends if you're a gamer amd sure. If you want like ultrabooks, intel still has a better package. By which I mean things like human presence detection. The idea that the computer detects you.
Leo Laporte [02:33:34]:
The stuff that's system on a chip is better.
Daniel Rubino [02:33:37]:
Yeah, intel and Qualcomm both do those. If you want 5G in your laptop, that's intel and Qualcomm. Qualcomm, okay.
Leo Laporte [02:33:43]:
So for mobile computing, AMD is not.
Daniel Rubino [02:33:45]:
If you want the latest technologies for a long time it's still hard to find an AMD ultrabook with a 4k touch OLED display.
Leo Laporte [02:33:54]:
Both my desktops though are AMD. Both my X86 desktops. Yeah.
Daniel Rubino [02:33:58]:
So it makes sense I think on desktop AMD are king, right. They do a lot of good there. But when it comes to laptops, and this isn't necessarily AMD's fault, it's just that intel has such a strong connections to the OEMs, they help design the laptops, they help with marketing and they offer these packages, you know, they do the WI Fi, they do the Bluetooth, they do the 5G, they do, they do all this and they can just offer a better package that's more affordable to these companies. So I don't want to tell this is good, right?
Leo Laporte [02:34:29]:
Because it means we have Snapdragon, we have x86, we have Apple Silicon, we have competition. That's good. And they all feed on each other. Improve on based on what the other Qualcomm's new chips.
Daniel Rubino [02:34:40]:
Yeah, Qualcomm's new chips have 80 tops for the NPU which is like that's.
Leo Laporte [02:34:44]:
Double what the original co pilot plus PCs did.
Daniel Rubino [02:34:47]:
And like intel will be at 50 for their new chips. So it's just like you know now what do you do with 80 tops? Well technically nothing. But the point is you, you put that hardware in people's hands, into developers hands and now you can do some pretty crazy localized AI. So it should be interesting.
Leo Laporte [02:35:01]:
Well that's why when you're running that.
Sam Abuelsamid [02:35:03]:
Localized AI, what that 34 hours of battery life. Was that become about 3.4 hours?
Daniel Rubino [02:35:10]:
No, actually. So NPUs are. That's like. What's awesome about the NPU is they're extremely efficient, so they can run tests in the background. So something like Windows Recall, which is basically taking thousands of snapshots of your computer is running on the MPU and it barely impacts battery life. NPU is like one of those things that is just like. It is so precisely designed for a specific task, but it's so efficient at what it does that it does. This is why Microsoft has been very stringent with the Copilot plus designation.
Daniel Rubino [02:35:42]:
Like desktops don't have it. Like I have a 5080, I think in this desktop. It's not a copilot plus PC. Why an Nvidia GPU with its tensor cores will crash crush any kind of NPU in a laptop, but because of the efficiency, because it's going to use a lot of power of that GPU just to do those basic tasks that it's not the kind of experience that, you know, Microsoft actually wants. So until they get NPUs into desktops, they're not going to allow, as far as I know, Nvidia to be able to do that task.
Leo Laporte [02:36:15]:
Well, I guess I'll just kind of keep an eye on things and just be happy with what I got.
Daniel Rubino [02:36:21]:
You'll be able to afford the new laptops and new Macs. I think you'll be okay.
Leo Laporte [02:36:25]:
I hope so. It's great to see you. Daniel Rubino is the editor in chief of Windows Central. That's where you can keep up with all of this, including the Dell and HP earnings from this week, which we didn't get to. But I think the PC Marketplace is warming up a little bit. It's very interesting right now. Although Dell said about half a billion people are not upgrading Windows 11, they're sticking with their old Windows 10 PCs, which cannot be good news for Microsoft. That's a little.
Leo Laporte [02:36:51]:
That's a little scary.
Daniel Rubino [02:36:53]:
Yeah, it's going to be. Yeah, no problem.
Leo Laporte [02:36:57]:
It's great to see you, Sam. Able Salmon, my car guy gives me great advice. In a year, you're going to have to help me decide what to replace the BMW with. We'll have some good choices by then, I think.
Sam Abuelsamid [02:37:09]:
Oh, yeah, yeah, there's, there's plenty of good options.
Leo Laporte [02:37:11]:
I just know I'm never buying a gas car again. I just love ev. Ev's. I'm so happy.
Sam Abuelsamid [02:37:16]:
Yeah, I mean, I think, you know, we've, we've come you know, now that my wife's been driving an EV since July, she has the Nero, right? Yeah. She doesn't. Yeah. She doesn't want to go back to a gas car.
Leo Laporte [02:37:26]:
It spoils you. Yeah. Now, if I had that hydrogen car, maybe. Is that. What is the torque on a hydrogen car? It's good.
Sam Abuelsamid [02:37:34]:
It's the same. It's an ev.
Leo Laporte [02:37:36]:
It's an ev, but it doesn't have batteries. It just burns hydrogen. Okay.
Sam Abuelsamid [02:37:39]:
Yeah. It's. It's generating the electricity on the fly instead.
Leo Laporte [02:37:42]:
Oh, I see.
Sam Abuelsamid [02:37:43]:
Storing.
Leo Laporte [02:37:43]:
So it still. It still has electric motors.
Sam Abuelsamid [02:37:45]:
Yeah. So fuel cell just takes hydrogen and oxygen and produces electricity and water you.
Leo Laporte [02:37:50]:
Could drink out of the tailpipe.
Sam Abuelsamid [02:37:52]:
Yes, but don't.
Mike Elgan [02:37:53]:
I. I drove to Los Angeles in my Prius, which is Plug In Hybrid, and my son. I was following my son because we were going to Universal Studios and stuff like that.
Leo Laporte [02:38:03]:
Ah.
Mike Elgan [02:38:04]:
And so I was following him, and, you know, you'd hit a hill and he hit the gas and you see this, like, water coming, just come spraying out the back. It was pretty cool. Look. Very different from a regular car.
Leo Laporte [02:38:16]:
Yeah. That's Mike Elgin. You can read about his AI journey at MachineSociety. AI. He also does a podcast with the wonderful Emily Forlini.
Mike Elgan [02:38:29]:
That's right. Called Super Intelligent.
Leo Laporte [02:38:31]:
I love the name.
Mike Elgan [02:38:32]:
We have a lot of fun. We have a lot of fun.
Leo Laporte [02:38:34]:
Yeah. And of course, gastronomad.net if you want to know about those incredible, incredible Grassdoor nomad adventures all over the world with Mike and Amira. We did one a couple of years ago in Oaxaca. It was one of the best travel experiences we've ever had.
Mike Elgan [02:38:49]:
Next year, Leo, is our 10th anniversary. Oh, my gosh. Anniversary. The founding of the company. And so we're going to be doing incredible. Like, we're gonna have a crazy, great year next year. So if you ever want.
Leo Laporte [02:39:03]:
All right. I think it's time for me to go on another one.
Daniel Rubino [02:39:05]:
One.
Mike Elgan [02:39:05]:
I think it is, too.
Leo Laporte [02:39:06]:
Yeah.
Mike Elgan [02:39:07]:
It's going to be fantastic. And travel gets worse and worse every year. It's gotten really bad lately, the crowds.
Leo Laporte [02:39:12]:
Thanks to Instagram. I really blame Instagram. It's all these influencers out there standing in the lavender and, you know, doing the tick tock dance.
Mike Elgan [02:39:21]:
Yeah.
Leo Laporte [02:39:22]:
And ruining it for the rest of us.
Mike Elgan [02:39:23]:
That's right.
Sam Abuelsamid [02:39:24]:
And just gotta go find a nice secluded beach somewhere.
Mike Elgan [02:39:27]:
Oh, we do. We do that. We find a lot. Nice, secluded. Be everything.
Leo Laporte [02:39:31]:
You find the places the influencers don't know about yet.
Mike Elgan [02:39:34]:
Exactly. And it's all. No Downsides, you don't have to rent a car, you don't have to deal with the hassles. You really want traffic.
Sam Abuelsamid [02:39:40]:
Do you curate the people who, who join your, your trips, you know, and make sure, you know, don't allow any influencers?
Mike Elgan [02:39:48]:
No, we have very. Actually by far Leo is the most influential person we've ever had. That is for certain. No, we actually, you know, we find it's very self selecting the people who do these experiences. Just fantastic people. It is, I don't know, it's a miracle.
Leo Laporte [02:40:03]:
But great conversations. You all sit around a big table. It's just a handful of couples, you know, maybe four or five couples. So it's a small group. Oh, I just really had a wonderful time. So what's the, so next year, huh? That's the, that's going to be incredible.
Mike Elgan [02:40:18]:
Next year is going to be really, really, it's going to be extra super special next year because we're celebrating the entire year or 10, 10 years doing this.
Leo Laporte [02:40:27]:
Congratulations.
Mike Elgan [02:40:28]:
Thank you. Thank you so much.
Leo Laporte [02:40:29]:
Yeah. 2006, it all began. Gastronomad.net and Kevin is still, I know he's working at Kagi now, but hello, Chatterbox is still going strong, right?
Mike Elgan [02:40:39]:
They have nice sales going. He's got other people running it. But, but yeah, it's teaching, teaching al literacy.
Leo Laporte [02:40:47]:
Smart speaker that teaches coding skills.
Mike Elgan [02:40:49]:
That's right. It's a, it's a great thing. People love it, schools love it. And, and so it's, it's still a going concern. And so many of the things that he developed, building this product around privacy around how to provide content that's really user driven. He's bringing some of that knowledge to Kagi, all of that knowledge to Kagi and it's, it's really a great fit at Kagi based on what he did here. But yeah, it's good. Chat box is still great.
Mike Elgan [02:41:17]:
Still a very good educational product.
Leo Laporte [02:41:20]:
Love to everybody, including Princess Squishy Face.
Mike Elgan [02:41:22]:
That's right.
Leo Laporte [02:41:23]:
Wished me a happy birthday and that was very nice of her. Thank you, Mike. Elgin. Thank you, Sam. Thank you, Daniel. Thanks to all of you who join us this week and every week we do Twit Sundays, 2 to 5pm Pacific, 5 to 8 Eastern, 2200 UTC. You can join us live in the club Twit Discord. Of course, we'd love to have you in the club if you're not a member.
Leo Laporte [02:41:44]:
Twit Club Twit. This is a good time to do it now through Christmas. There is a 10% off coupon so a good time to gift yourself or someone in your life with the gift of Twit. We also stream it live for everybody. YouTube, Twitch, X.com, facebook, LinkedIn and Kik. So you can watch us live at that time. But really most people prefer to watch at their leisure at their convenience, which I would encourage. Download a copy of the show, audio or video or Both from Twitter TV the website, or there's a YouTube channel with a video dedicated to Twitter.
Leo Laporte [02:42:21]:
Great way to share clips or subscribe in your favorite podcast client. That way you'll get it automatically the minute it's available. I think we were like number 92 on Apple's top podcast lists. We were somewhat far down, which is true because, you know, you've been around for 20 years, you don't get lot of new subscribers every week. But help us, help us a little bit. Put a nice review there on the Apple podcast app or whatever app you use. Let the world know about the probably longest running tech podcast out there. Certainly one of them.
Leo Laporte [02:42:55]:
Anyway, this week in Tech. Thanks for being here everybody. We'll see you next time. And as I've said for 20 years, another twit is in the can. Take care.
