00:00Let's start with AI because, you know, you think about AI and it being infused in everything.
00:06I mean, you and Romain were having an interesting tangent before this program started that you think about the state of AI when it comes to hardware that we use.
00:15You think about phones and everything else.
00:17And I just wonder where that's going.
00:20And in your view, I mean, what sort of do you think is the path to having it actually be a bigger part of our lives beyond just chat box?
00:28Sure. Well, back in 2011, when we actually introduced the Nest thermostat, we had AI in it.
00:33We just couldn't say it. So we've AI has been around the world for a long time.
00:38It just became a big thing about three years ago. And now we're seeing really AI in the data centers and much more powerful models.
00:43It's now moving back to the handsets. It's going to be coming.
00:48It's still not there yet. But hopefully Apple is going to be able to do that because they actually have AI engines, hardware engines on the hardware if they get the software there.
00:56So I think it's going to it's starting to happen because it has to happen because it's much cheaper to do the compute locally where they don't have to build more data centers and they can make it available and much more quick, quick, quick to operate.
01:08It's a great point that you think about, you know, what we talk about as AI today.
01:12A lot of this is rebranding. To your point, AI has been around intangible for a while now.
01:17But when you think again about, you know, it being in some of the devices that are glued to our hands in a lot of cases, are you surprised that it's not more ubiquitous at this point?
01:26No, actually, I'm not, because what's happening is we have these bigger is better in terms of the models.
01:32And that's not actually what the consumer needs. They don't need to go and do chemistry analysis on their phones.
01:38They need to do simple things.
01:39And so what we're what the popular media and where the business is all investing in these really big models of this stuff for consumers to actually use the use these things, it can be much smaller.
01:50But the attention, the product details, all those things haven't gotten that because everyone's chasing these bigger data data center centric models.
01:59And so we're again, we're going to we're going to see this coming in the next couple of years.
02:03Is the approach that we see here in the with a lot of the U.S. based companies, is that the same approach that you're seeing, say, in Asia, specifically in China and to a smaller extent, Europe?
02:13Well, that that's a very interesting question.
02:15So what you're seeing is a lot of small models that are almost as capable as the very large models coming out of China.
02:21Why? Because they don't have the compute resources.
02:24So I always love resource constrained environments because it brings up more creativity.
02:28And so that stuff's happening. And you're seeing a lot of developers in Europe and the U.S. using those models because they're open source because they're cheaper to work with than the than the closed ones now.
02:39Well, I want to go there, too, because we look at all the cash being splashed in the space and we look at the giants in this space, the open AIs, anthropics, et cetera, as well.
02:47And I almost wonder if that almost handicaps them in a way, because you think back to sort of the last big tech boom and just, you know, I know there's a lot of mythology.
02:53And, you know, you know, Bill Gates in his garage or maybe not his garage or Steve Jobs and all these people.
02:59But there was sort of this underdog mentality to a lot of those folks that they were sort of fighting the big man.
03:04And they had to be innovative because nobody really took them seriously.
03:09And I do wonder if if, like I said, the people now at the forefront, the Sam Altman's and the Jensen Wong's are to a certain extent.
03:16I know it seems odd to say they're at a disadvantage given how much money they've raked in.
03:19But I do wonder if that puts them at a slight disadvantage.
03:21Well, first of all, I would put Jensen and Sam in two very different camps.
03:25Jensen is really innovating.
03:27I think Sam is doing a lot of spraying and praying.
03:29He's getting a lot of capital and trying to find an application model where he can actually get success for a revenue thing, kind of like Anthropic.
03:37Anthropic is actually getting really tight.
03:39Their application is really around coding for the first thing.
03:43And they're winning.
03:44And they're winning definitively.
03:46Open A is still a platform.
03:48It's not an application.
03:49And the best companies are application companies, not platform companies.
03:54The last really great platform we've had was Microsoft Windows.
03:57There's been nothing since that.
03:59So you always have to target the application.
04:01Well, speaking of spraying and praying, I'm so curious to hear your thoughts on what's going on when it comes to the tech talent wars, if you want to call them that.
04:08Because some of the packages that we're seeing handed out are just mind-boggling in terms of the dollar signs there.
04:15I mean, think about that allocation.
04:17Do you think that, you know, that is money being spent wisely?
04:20Or are we at a place where it's just who can command the most and where?
04:24I think, you know, we saw a little of this in the early 2000s with Google and then Facebook when Almeida, when they were trying to get at talent.
04:34Now it's really gone crazy.
04:35And I wish I could say that, you know, it makes any rational sense.
04:41It doesn't.
04:42I would love to be an engineer back then.
04:43You know, it's like, wow, you don't have to be an executive and wait 20 years to get those kind of packages.
04:48But I think that's just a moment in time.
04:50I think that's going to all get readjusted.
04:52And maybe some of those companies actually can't pay those salaries long term.
04:55They're going to have to re-contract them.
04:58I do want to just go back to the hardware side of it, given your background, and also given the idea of kind of the maturity of the iPhone and, for that matter, the maturity of the Android phones.
05:06And kind of where you think the next iteration of devices is.
05:10Is it going to be even a phone?
05:12And if so, I mean, what does that look like?
05:14And who do you think is actually maybe making progress towards actually introducing that?
05:18Well, the thing is, you know, we have so many devices.
05:21We have a plethora of devices, watches, rings, you know, glasses now.
05:26So that kind of suite of things on your person has been really filled out.
05:29And so now it's about software models that actually takes advantage of all that input and output, the data and the sensing and the output, to be able to use it wisely, to be more like the movie Her, you know, and trying to make that happen.
05:42But that takes a ton of software, and that takes a ton of software, not just the LLMs we're hearing about, but other types of AI to make this all work and work together.
05:52All those devices should work together.
05:53Does Apple have that, or is it going to be a software company out there that ends up maybe leapfrogging the giants of today?
06:00I think only if you have enough of that hardware suite, you know, the rings, the whatever it is, can you get enough data to inform the AIs to get better and better,
06:10to allow them to put back onto these devices?
06:13Because it's not just one mode of one device that you use.
06:17It depends on the mode of use.
06:19So if you're on a plane, a train, in a bus, walking, and you're in office, and you're in home,
06:24it's a very different context and a different type of device you might be using and a different, you know, set of data that you're collecting.
06:31So you really, it's not going to be a one-size-fits-all.
06:34The smartphone is incredible, and the thing to understand is that the best way to visualize visual data is with a screen.
06:43It's not through your ears.
06:45And so, yeah, you'll say, oh, we have AirPods, and now we can do a bunch of voice.
06:49But when you don't want it to describe a map, you're going to bring out your phone, and you're going to look at it.
06:54So it's more like a, you know, the old keyboard and mouse, and it's going to be like that.
07:00Now, obviously, the applications are going to get better and much more voice-centric.
07:04But we have to wait for all of that, and it's not just necessarily a new device.
07:09And, Tony, before we let you go, you said it was recently as July that when it comes to Apple, they need to get bolder when it comes to AI.
07:16It's only been a couple of months since then, but you think about the last few months of progress.
07:20Have you seen anything that would make you feel a little bit more positive there?
07:25Well, first thing, they did a big deal with Google, right, last week, I think it was.
07:29And so that's a great first step.
07:32I'm hearing bubblings and rumors, you know, smart cameras coming from Apple for the home, various other kinds of smart display in the home.
07:41So I think it's coming.
07:42It's bubbling up, but it takes a while to, you don't just turn those companies on a dime.
07:48So we'll see, hopefully.
07:49Do you still talk to people at Apple?
07:51Absolutely.
07:51I was talking to some people this morning.
07:53Okay.
07:53All right.
07:53Fair enough.
07:54You want to tell us what it was about?
07:56We were talking about Vision Pro.
Comments