Tokens Are the Next Commodity

00:00:00:00 - 00:00:08:19
Unknown
This episode of the AI Inside podcast is brought to you by Airia. Get started for free today at airia.com.

00:00:08:21 - 00:00:33:06
Unknown
Coming up, Jeff Jarvis and I were following Nvidia's GTC event and Jeff followed it very closely. He's going to detail the company's Viera processors, Nemo claw and the dLSS five update that brings photorealism to its GPUs. Plus, more meta layoffs. OpenAI has focus on its core business, whatever that means. The gen AI resurrection of Val Kilmer and the insignificance of Encyclopedia Britannica.

00:00:33:06 - 00:00:46:01
Unknown
All that coming up next.

00:00:46:03 - 00:01:04:18
Unknown
Hey everybody, welcome to the AI Inside Podcast is this show where we take a look at the AI that is layered throughout the world of technology, like a fried egg sandwich. She got a little mayo. You got a little salt and pepper, some fried egg, maybe a layer of cheese. Oh, sorry, I, I didn't eat enough for breakfast.

00:01:04:18 - 00:01:21:21
Unknown
Apparently, I'm, one of your host, Jason Hall, joined, as always, by Jeff Jarvis, who I'm sure ate better this morning than I did. Well, yesterday, I actually had something layered that was pretty disgusting. Oh, so what did you have? You could be a metaphor for parts of overdone I, I sure I tried, I didn't need all of it.

00:01:21:21 - 00:01:48:19
Unknown
About a third of the McDonald's arch. Big arch. So I mean, what's the deal with that? It's just a really big burger, right? It's a really big burger. And the president of the company, supposed do a video supposedly eating a particular devious little bite. Yeah. In truth. So it's two and a quarter pound. So it's a half pound of beef, three slices of cheese, lettuce of, crinkled onions, and of course, a new special sauce.

00:01:48:21 - 00:02:07:06
Unknown
And the problem with it being in layers, since we're talking about layers here, is, as you in fact, try to eat it. The special sauce is so slippery that the beef patties start slipping out. Oh, I hate that. Yeah, I hate that unbalanced burger. No no no no no no no no. That's why you that's why the cheese is there the cheese.

00:02:07:06 - 00:02:22:03
Unknown
This is the is the cement. It holds it together. Exactly. So I ate about a third of because I was not going to eat at all. I mean I actually, but I'm anemic right now, so I need beef. That was my that was my right here. And I can say that I tried the bottom was pretty good, I'll give it that.

00:02:22:06 - 00:02:40:00
Unknown
But that was just disgusting. I didn't even take a picture of it. I it was to have appetizing this disgusting mess left on the box. So yes, I can sometimes overdo it like a big arch. Like it like a burger. It can be. Oh, yeah. Yeah. The layer, the layers just don't hold together. And they lose out.

00:02:40:02 - 00:03:02:09
Unknown
Yeah. Well, I, I, normally I don't eat McDonald's very often, but I was traveling this week. I was in Saint, San Antonio for a cool event called Sponsor games. And, at the airport there was a McDonald's. So I went there for for my, dinner on the way back, I was in a hurry. Yeah. And I saw that on the menu.

00:03:02:09 - 00:03:21:16
Unknown
I was like, no, I can't do it. I just I can't do it. It's too much. It's ten bucks too. It's I love box. I mean, fast food isn't cheap anymore. It used to be the cheap solution, and now it's. Nothing's cheap. That's the thing. Nothing is cheap. Jeff. Well, it's good to see you, but. But it's free to watch us, folks, so.

00:03:21:16 - 00:03:44:19
Unknown
That's right. That's right. Yes, yes. You choose to become a sponsor on on Patreon, but you'll hear about that momentarily. That's voluntary. That's that's true. It's totally up to you and we appreciate it deeply. And who knows, maybe we'll start sending you big arches as our thank you gift. I suspect it doesn't go bad. That's the preserve. Yeah, we won't actually do that.

00:03:44:19 - 00:04:03:12
Unknown
That's a horrible idea. But apologies for a day late show. Like I said, it was traveling. I didn't get back until, like, midnight last night, so it was a late night. So we couldn't do the show yesterday. We have it rolling today. Everybody who's watching the live stream. Thank you for the the kind of, stop, start.

00:04:03:12 - 00:04:26:20
Unknown
Because I think that because I was in Central Time Zone when I scheduled it, it ended up booking it two hours earlier. Anyway, to peel back the. We're peeling back the layers of this live show, probably more than most people care about, but. So that's what happened this morning. Nonetheless. We're here and we're ready to talk. It looks like we're ready to talk about Nvidia in this top block because like I said, I was at an event.

00:04:26:23 - 00:04:49:13
Unknown
I didn't get a chance to watch all of the announcements in real time, the way I know you did. Jeff, you love follow what? I am a connoisseur of Jensen Wong's keynotes. I watched the ball. I'm just amazed by them. What the showmanship. But that's not really it. Because it's not flashy. It's just impressive. He knows his stuff.

00:04:49:16 - 00:05:13:20
Unknown
He has a command of a very complicated, topic. Of course. He's in charge of the company. He should not too many extra points for that. But it's really impressive how he can, explain things clearly. He's a hell of a communicator. The interesting about this keynote was for the first hour, it was all context. It was all, you know what?

00:05:13:23 - 00:05:33:19
Unknown
Everything that Nvidia does, everything gets done. And I thought, well, there's nothing new here. Yeah, there's a bit strange. And I thought, well, maybe. Does it go ahead. Does does the format of, of GTC kind of follow other like let's say developer conference. Right. Like I think of of Google IO where they've got the big like 1 to 2 hour keynote at the top.

00:05:33:21 - 00:05:54:21
Unknown
And then there's all these smaller events that happen. Oh yes. It, it's I think it's three days long. There were tons of sessions, tons of people there. All kinds of tracks. But the keynote supposed to bring it all together and unlike, let's say, iOS keynote, which we've both been to in the scene, where, they come out and they bring out all kinds of people and it's all prepped and it's bam, bam, bam, bam, bam.

00:05:54:21 - 00:06:23:27
Unknown
Yeah. And you're and you're watching things, like it's out of a of a radio carousel. Yeah. Right. At Nvidia it is Jensen Wang show period. Full stop. Occasionally he did it this time. Occasionally he'll bring out somebody from another company to say hi and then get him off the stage. He'll inevitably bring out a robot or some equivalent, which he didn't time, to share the stage with, which always doesn't work fully.

00:06:24:00 - 00:06:47:17
Unknown
But it's his show. It's completely his show. Shows over the first hour, he gave context and background and what I like about his keynotes, too, in general. And he didn't so much do at this time, is that each one has a different lesson that he integrates into it. It's explaining scale or, or, tokens, the tokenization of the economy or the digital twin, and so on and so on.

00:06:47:20 - 00:07:14:09
Unknown
This time I think he wanted to explain Nvidia because what came at the end then, was an important strategic, I wouldn't say shift, but revelation in the beginning, he wowed the audience and the market when he said, and I want to get this right. I hope that last year he had $500 billion revenue, and he predicted that in 2027 it would be $1 trillion revenue.

00:07:14:12 - 00:07:36:06
Unknown
And the stock went down again. But but it was momentarily, of course, excitedly up react. And so that was news. That was big news that that he was he was saying that. But as he got toward the end, you'll remember that they did a weird deal, not an acquisition, not an acquire, kind of a licensing, but kind of a higher but kind of an investment in grok.

00:07:36:06 - 00:08:00:18
Unknown
Not that grok. Grok with a Q right. But you right. Is a is a chip. That's that was kind of missing from the Nvidia family. And it starred in the show here because it is the essence of inference. And, what was fascinating about the whole show was that at the end, Wong said that this is the age of inference.

00:08:00:20 - 00:08:23:14
Unknown
So let's remember that, the Grace Blackwell, the hopper, the Vera Rubin chips are the ones that are that are lusted after for training models, and they're still in huge demand and still can't hold up the demand. And there were some stories saying, well, was, Nvidia going to get stuck in the in the age of trading? Because grok has a chip that enables it.

00:08:23:14 - 00:08:51:05
Unknown
I can't explain it fully, but but what I understand about it when they did the deal originally is that it enables, greater memory access for the process of inference and that speed and that access is so important. And so he's, co-branded the chip offering the chip out there, offering inference, saying the inference is there. And then out of that, he went open for crazy.

00:08:51:08 - 00:09:07:02
Unknown
Really talked a little bit about that last week. He drove to the deep end. He dove in the deep end. And there was we kept waiting for it to happen waiting for it to happen. Then at the end he really went into it. So they, they've created I'm jumping all around a run down here, I apologize.

00:09:07:06 - 00:09:29:19
Unknown
Oh, no, no, no, it's good. I just, he created, Open Claw, which is basically, I'm sorry, old Claw, which is open claw with guardrails. Right. It's it's it's it's, enterprise ready. But he sees a world here very clearly where people are going to create like crazy. They're going to do it within companies. He said something very interesting.

00:09:29:19 - 00:09:47:24
Unknown
At one point he talked about compensation to employees, and he said, you know, I want to offer them, compensation money in stock, but I also want to offer them tokens. Yeah. What are the next commodity is, is something that he said or the next tax free compensation. Right. Yeah. And I don't know what he intended that to be.

00:09:47:24 - 00:10:04:14
Unknown
Clearly, if you're an employee at Nvidia you're building something. You should be it all the tokens you need. Because you're an Nvidia you should do what it is. So is this so people can build their own companies, build their own things in their own lives? I'm not exactly sure, but he's one of his one of his sessions, earlier sessions was about the economic value of tokens.

00:10:04:14 - 00:10:29:25
Unknown
Tokens is the key to this new economy. Can I can I jump forward for just a quick second? Because maybe, maybe this is part of it? And, let me see if I can pull that up because it's a little bit further in the rundowns, but I'm trying to move it up as we go. There was this talk with Sam Altman where he says, we see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter.

00:10:29:27 - 00:11:09:14
Unknown
This is separate. This is just a separate talk. But I, I'm wondering if maybe that's a little bit of the direction that we're going here, which is like tokens, the foundation of so much of what we envision the future to be in technology. It becomes a utility. I think the contrast is great. I'm glad you brought that up, because I think that that one said it far better that all but one said there is this, this, this, commodity, this this, widget, this, this entity, economic entity that powers AI.

00:11:09:17 - 00:11:30:04
Unknown
And it's a tokenization of everything. And he again, he's talked about this in the past. He really thinks this way, where it's all been started, I think very obnoxiously, even odious Lee saying that intelligence is something he can own and he can turn a tap on and off on something that he calls intelligence and, and we're going to come back to that topic in a few minutes.

00:11:30:04 - 00:11:58:03
Unknown
What I plug my new book series, because with one of the first authors talks about this, there that this notion of intelligence and what intelligence is, is so important, what all been managed to do in one fell swoop there was to eliminate humanity from intelligence. He eliminated learning and, education and thinking and philosophy. And he acted as if intelligence is his commodity that he can own and control and turn on and turn off.

00:11:58:06 - 00:12:23:25
Unknown
Yes. That's right. One doesn't say that he says something that's far more tactical that there is this, you know, electron of of of information now electron of computation, which is the token. And interestingly, you know, I thought about this as yesterday, they make money. The models make money both ways. The tokens you put in with your prompt, you pay for the tokens.

00:12:23:25 - 00:12:43:07
Unknown
You get out with the response. You pay for. You pay both ways. Meanwhile, you maybe treat teaching that maybe to other things, but so so the token becomes like, maybe I like that better. The token is kind of an electron. It it is the thing that traverses this communication. Whereas Albert, I think I said it quite obnoxiously so.

00:12:43:09 - 00:13:14:27
Unknown
Among other things, there was also in the talk, Grace Blackwell, which is the current top chip. There'll be a next one coming soon, which he's talked about. The Feynman chip. But Grace Blackwell is wildly powerful. Now, what's available in desktop computers with Bondo desktop computers, computers that you and I can't afford. But this fits in with the Nemo claw, because the advantage and the and the and the security necessity of open floor is you need to run it locally.

00:13:15:00 - 00:13:34:11
Unknown
Because if you set this thing loose on your company, watch out. As we saw a few weeks ago, the head of, was the safety and security at meta. Her agent erased all of her email. Yes. You don't want to let it loose. You don't want to lose your credit card. You don't want to do certain things, but you do want to be able to create these agents.

00:13:34:13 - 00:14:01:29
Unknown
And it's safest if you run them locally. So to be able to get a a Grace Blackwell chip on your desk is really important. So they delivered the first Dell version of this. There's various OEMs to Andre Karpathy who also interestingly last week I don't think it's on the rundown. He put out the smallest possible version of a model so that you could run it on your own.

00:14:01:29 - 00:14:22:06
Unknown
You could learn with it, you could train it on your own and see what it is. So I think we see a lot of what's happening here is, on the one hand, Nvidia is supporting the growth of these gigantic data centers. And it's a very a silicon that's at the heart of of all of those. And that's always going to be necessary for the super scalers.

00:14:22:08 - 00:14:44:28
Unknown
But on the other end of this, we start to see a consumer end, that is powered by this through open claw, through animal claw, through those kinds of things, which I think is is fascinating. One more thing then I'll then I'll show you probably probably a lie. The other thing occurred to me after watching the Jason was that in my world, in media and our world in media, right.

00:14:45:01 - 00:15:03:14
Unknown
You had to have access to the premium press, the broadcast tower. You had to have capital. You have to have tools. You have to have expertise. And obviously what the internet did and the tools that are the rod thereupon make it possible for us all to create, ergo what we're doing right now. Right. Not true. With technology in terms of computing.

00:15:03:17 - 00:15:26:12
Unknown
Really. Yes. You could write a little basic program, you could do this, you could do that. But I seem to be this wildly, hugely powerful thing that you couldn't touch unless you had billions of investment and billions worth of chips. And now I think we see the possibility of this being democratized to the extent you can run something locally.

00:15:26:14 - 00:15:44:01
Unknown
And the guessing is that all laptops are going to get more expensive because all laptops are going to want to run this stuff. And maybe this widely leaves me out with my Chromebook, because it doesn't run stuff. And, so, so everybody can have these things that they can run locally and they can just tell it without going to terminal mode.

00:15:44:04 - 00:16:04:16
Unknown
That's what I'm waiting for, that that's the next big step here is moving past the need for terminal load, where you can just say, I want you to do this and it programs whatever it needs to. And it does. It. And I think that's we're definitely moving in that direction are moving the direction in that direction. They're already kind of flirting with that, that sort of integration as well.

00:16:04:18 - 00:16:25:13
Unknown
So what fascinated me in the end maybe not for all of the things you're talking about. Oh yeah, definitely. But this is what, you know, Jensen Huang you know, it always gets me, gets wheels going. And then the one hand, he his is the ultimate business. Business company, right? He is serving gigantic companies, with the highest end technology we have today.

00:16:25:16 - 00:16:45:21
Unknown
But on the other hand, suddenly there's a now a new retail possibility, a new B2C company. Right now you got to have, I don't know, 5 or $6000 to buy a great stackable chip on your desk. You got to know what to do with it. And, this could be like the early days. Somebody said one of the comments on all this.

00:16:45:27 - 00:17:15:12
Unknown
Yeah. For years and years and years we were all gonna be running Linux and you know mortals are running Linux. Just geeks running Linux. Right. And this could be the same way. But the difference here is you don't have to code. And if you can avoid terminal mode and the freight that comes with that for most mortals. And if you can have this power on your desktop to again, with just language, say, this is what I want you to do, and it does it that I see why Wong went kind of crazy at the end about this.

00:17:15:12 - 00:17:59:08
Unknown
So he went about two hours, 15 minutes. Yeah, that was going to be my next question. He usually has a pretty chunky long keynote and incredibly dense. Obviously there's yes, he started late. But you know, God knows somebody backstage was getting eviscerated for a slide. Not working. Sure. Probably so. But he went long. The other thing I say, I was wrong, I lied there for more than the other argument he was making in the first hour was that it was very interesting that that if you build a, you know, in gigawatt, super watt, whatever, data center, it's limited by its power and you can't really upgrade that.

00:17:59:11 - 00:18:30:06
Unknown
But what you can do is get faster. And when he talked about the sequence from the hopper chip to the, Blackwell and Vera Rubin chips, he showed how the speed and efficiency of what, Nvidia offers and how it upgrades you all along the way with, its operating system and with its tools. And that creates a huge opportunity.

00:18:30:09 - 00:18:51:23
Unknown
And he put forward a business model that says that you're going to have some tokens that are free always. You know, you go to open air, you go to Gemini, you're going to do that, but then you're going to see sequences of expense that the market will bear to pay a lot more for those tokens if it can do speed and power and so on, so forth.

00:18:51:26 - 00:19:11:01
Unknown
And that what he's arguing is that your cost of delivering that is going to go down, is going down as the chip's power goes up. Right. So he wanted to kind of explain all that and set the stage. I feel like I'm trying to summarize a philosophy talk because that's what is so fascinating about him. And and again, at the end, you there's a great moments.

00:19:11:01 - 00:19:27:24
Unknown
We're up from the stage. Come the chips in the in the racks and and the thing is to show and we don't know what we're looking at. Nobody is of those we're looking at. But everybody's always at all of these things because it makes it tangible and real. But what what the exciting stuff isn't really the chips. It's what we're going to do with them.

00:19:27:27 - 00:19:55:11
Unknown
Yeah, yeah, yeah. Thank you. That's a fantastic summary. Of all of that, almost all of of their announcements. But, I think one thing that stood out for me kind of reading through and everything is just that Nvidia has done a really great job in the past handful of years, making a strong case around its hardware, around its, it's, you know, it's accelerators and GPUs and all that.

00:19:55:13 - 00:20:23:02
Unknown
But now with that claw, you know, with their Nemo claw and everything, they're kind of also making the case for, for the software side. And, you know, maybe, maybe that ends up being part of that bridge that you're talking about to the consumer. A I think one thing I came across as a genetic, as a service, and then when I put that into the acronym is, it's basically asked which could get you in trouble, but hey, there we go.

00:20:23:02 - 00:20:41:05
Unknown
It's an acronym. What are you going to do? But you know, speaking of the consumer side, this is really interesting. Because because probably before anything, because I was really off the grid, I was really focused at my event, not on tech news at all. On everything that I was. Money makes the world go around the world.

00:20:41:12 - 00:21:03:09
Unknown
You know, world content is driven by sponsorships. So learning how to do it. So, does work. So, so, you know, if you continue to hear, sponsors on this podcast, you'll know that it was money well spent on my hand and time well spent. But anyways, I did here while I was there and checking minimally on news reaction to one of the announcements that happened during the event.

00:21:03:09 - 00:21:34:08
Unknown
And I'm sure, Jensen showed this off, you know, on the keynote, which is dLSS five, which is a neural lighting system on RTX 50 series GPUs. Now, GPUs, we're so used to thinking now in a very short amount of time about GPUs in terms of AI, but, you know, Nvidia's bread and butter prior to this AI moment and and repurposing GPUs towards I was about kind of graphics, you know, I mean, it's a graphics processor processing unit.

00:21:34:10 - 00:22:04:23
Unknown
And so a lot of it had to do with gaming and, and all that kind of stuff. And so this particular announcement kind of ties back into that, but it's not getting a whole lot of positive reaction from the people who would really who you would imagine would really appreciate it. The, dLSS five is essentially aiming to improve computer graphics by upgrading by upgrades, raising, basically making computer graphics into photo realistic, resolution.

00:22:04:23 - 00:22:24:23
Unknown
So, you know, instead of kind of leveraging higher frame rates in, let's say, a video game that you're playing, it would focus its efforts on uprising, the actual imagery that you see. And if you're watching the video version, you can kind of see on the left is, let's say, a more computer generated sort of, sort of thing that we might be used to.

00:22:24:23 - 00:22:47:13
Unknown
And on the right would be the more photorealistic interpretation that that hardware is apparently capable of doing, kind of doing this on the fly up upgrade to photorealism, thanks to generative AI, Nvidia says that this is the biggest breakthrough in computer graphics since, the debut of ray tracing in 2018. That has been a huge deal for gaming.

00:22:47:15 - 00:23:08:25
Unknown
But I definitely saw a lot of very pointed reaction, to it. A lot of, a lot of people just kind of saying like, no, this is this is absolutely awful. This is basically, do I not have the ability to swipe it back and forth on this now, apparently I don't, that this is basically AI slop in video game form.

00:23:08:27 - 00:23:26:26
Unknown
And, I don't know, it's it's an interesting, conundrum. Thankfully, you will be able to turn it on and turn it off if you don't care about it. And you actually beyond that, if you just absolutely hate it, which I know a lot of people do, you could turn it off. But, it's interesting where we're at with this AI slop.

00:23:26:28 - 00:23:50:05
Unknown
Debate, because when I look at these two things, I'm like, well, this does look more defined than that. But I think because people know that it was oppressed by AI, they're like, no, I don't want it. Gamers can be so damn persnickety. Yeah. You know, they're passionate about it. Yeah. But I, you know, as as I looked at the images and I think, it's about artistic control too.

00:23:50:05 - 00:24:09:04
Unknown
And as a machine to do. Yes, I would do it all that. That's fine. Fine, fine. But, you know, for someone who cares about making things, it makes it better. It's just clearly, esthetically and realistically better. And I can do it at a decent computing rate. So it's very impressive. I don't think it's anything to lose sleep over.

00:24:09:06 - 00:24:27:15
Unknown
I think it's, it's it's, just one more advance in how, people are going to create on these tools with better and better stuff. Yeah. You know, so I pulled up my notes. I forgot I should do that earlier. Of course, we talked about, just a couple more things. If I can go back before I leave.

00:24:27:15 - 00:24:54:03
Unknown
Yeah, please. He emphasized how Open Claw was the most instantly popular open source project ever. And I mean this of the, you know, the uptake of Linux and so on and so forth. So I think that's that's true. He said. Companies are going to have to be asking what's your open claw strategy? And he really embraced this in a way that that surprised me.

00:24:54:05 - 00:25:29:24
Unknown
This is this is as big a deal as HTML. This is a as big a deal as Linux. Do you agree with it? Wow. Wow. I, I would, I would say that it's a yeah, I it's it's hard to say I agree with that because it's so new, but I mean, what it, it purports to do, what it can do as it, as it develops out, just kind of the expandability and flexibility of it and what it, what it I imagine would be I haven't installed it on my machine.

00:25:29:24 - 00:26:09:26
Unknown
Like, I'm still a little hesitant to install on my machine because I don't want to put everything into a sudden, you know, bad state and risk losing anything. I know what this about myself. I'm I'm I'm smart enough to, like, totally be dangerous with my stuff, you know, not smart enough to avoid that. But I think that what it promises and what I've heard about it, what people have been able to do to it, people who are much braver than I, kind of sets a scene for a really interesting future where these things are our true agents that can command these things in a way that our current kind of a genetic infrastructure, sometimes

00:26:09:26 - 00:26:30:19
Unknown
promises, but doesn't deliver on. I don't know, I, I don't I do feel like it's overstating it, but at the same time, I, I understand that it's a pretty darn big deal. And that's why people are responding to it. So, so wildly. Yeah. Yeah, yeah. So Nvidia was a big deal. I don't know what that stock's doing right now.

00:26:30:21 - 00:26:54:12
Unknown
See, Nvda, that's down a little bit. The market didn't go crazy for it, but I think there was even with a trillion revenue. Which was an odd promise for a CEO to make. I'm guessing a lawyer somewhere in the back was going, you sure you want to do that? You're sure you want to go there?

00:26:54:14 - 00:27:11:25
Unknown
But, but it was, as always, just fascinating. Yeah, yeah. And, I don't know, it might be done now. GTC I mean, that was Monday, right? 16 now I'm sure the event is probably winding down. I know a bunch of our, you know, a few of our friends were there. If I was in town, I probably would have been.

00:27:11:26 - 00:27:31:14
Unknown
Yeah, I would love to at least go for it for a day, to check it out, but I had to miss it this time around. Maybe next time. Thank you for watching and for summarizing. That was better than any of the notes that I took, that's for sure. Also, big thank you to our amazing patrons.

00:27:31:14 - 00:27:51:16
Unknown
Patreon.com slash I inside show you all make the world go spin around on this show. And we can't thank you enough for that. Thank you to Chris Houston and Daily Tech news show. Yes. DNS is, is a patron of our Patreon, so we appreciate them. Spreading the love a little bit.

00:27:51:19 - 00:27:57:26
Unknown
I also want to take a moment and thank the, sponsor of this episode that is.

00:27:57:26 - 00:28:20:06
Unknown
Area. Areas. Fantastic area is an a genetic platform that, just allows you to get really granular and set up all of these agents and tools, you know, bring in your MCP tools and create these, these, this architecture that is just really, really super compelling. The more you use it, the more your, your imagination runs wild.

00:28:20:06 - 00:28:42:29
Unknown
So it's definitely worth checking out. And I shouldn't be limited to just your most technical team members with area, everyone gets a seat at the table. So you've got your business teams, you got your operations, your engineering. Everybody gets to, you know, gets to check out these tools, the no code, low code and pro code approaches of the platform.

00:28:42:29 - 00:29:06:11
Unknown
So it really does make it very easy for employees at any skill level, right. If you're just learning how all this works all the way up to being super proficient, you can just do some wicked crazy cool things and start using AI in, very practical, very meaningful ways. And that, you know, that means, like your marketing team, they can automate their workflows.

00:29:06:11 - 00:29:32:01
Unknown
If you have analysts at your company, they can move faster doing what they do. Your developers can go even deeper than you're going to be able to. Of course, that's the power. They're all inside a single environment that's built for collaboration and security, built for sharing across teams and all that stuff. So instead of AI living off in silos area, helps you spread it across your organization in a way that's manageable.

00:29:32:01 - 00:29:52:08
Unknown
And, even the people working in my garage and, doing some construction, agree. They like it, too, apparently, if you hear them. My apologies. More innovation, more productivity and fewer roadblocks. To get started. So if you want AI to work for your whole team, not just a handful of specialists, area is built exactly for that.

00:29:52:08 - 00:30:23:27
Unknown
And you should really check it out for yourself. Learn more and get started for free today@area.com. And I'll spell that for you because it's not as it sounds. It's a AI RPA area.com. And check it out. You can kind of start, you know, getting in the sandbox and seeing how how it all works and start to get a sense of how, you know, the agent ecosystem and all the that the, the building blocks and all of the different models that they have available.

00:30:24:00 - 00:30:41:23
Unknown
It's it just makes a lot of sense once you're in there. It's very visual. I'm a visual person. It's really great. So area.com and we thank them for their support of the AI inside podcast. We're going to take a break and we'll come back and we'll we got a bunch of cool stuff to talk about, a little meta, little OpenAI, a little Val Kilmer.

00:30:41:23 - 00:30:46:29
Unknown
You probably didn't see that one coming, but it's coming up here in a moment.

00:30:47:02 - 00:31:10:25
Unknown
All right. So we got meta. I feel like this is a story that seems to come up pretty often these days, planning to lay off another 20% of its workforce. 79,000 person workforce, 20%. That's a lot of people. Is this the year of efficiency, remember? Continued. I feel like, wasn't that 2022 that that's that was a long time ago.

00:31:10:25 - 00:31:38:21
Unknown
But I feel like we're still in the year of efficiency for Matt other. So let me find some numbers here. So Benedict Evans did some calculations they said made us better end 2019 that is to say pre-pandemic, with 45,000 employees hired 27,000 across 20 2021 and 2022. Did it 11 K I expect you all to do the math by the end of this, 11 K layoff in 2023, but still end of the year up by about a 15 K.

00:31:38:23 - 00:32:11:26
Unknown
Okay. You got there so far. Cut another 20 K in 2023. Added 12 K back by the end of 2025 bringing it to 79 K. We started the story with 45 K. Okay. And if it now cuts 20%, that's about 16 K. That'll bring it back to the mid 2020 numbers. Okay. All right. It would also means has been that meta have laid off more people since 2019 than the company actually employed at that.

00:32:11:29 - 00:32:37:15
Unknown
Oh boy. Wow. Everyone interesting compares. He concludes everyone over hired in the pandemic. But this seems careless. Yeah. Yeah. I mean, you look at that, Jack Dorsey, you know, getting rid of what was it to 20%, 40% of his company? But he 40 hire. Yeah, he did it over hire as well. And certainly meta over hired went nuts.

00:32:37:17 - 00:33:06:08
Unknown
No fun to be one of those 20% of people. I mean, and a lot of I mean, that's a lot of people. It's it's crazy, the shuffle. I think there's major desperation about, I think it kind of feels that way. Yeah, it feels very, very sporadic and very, you know, stop, go. And I mean, you know, just days ago, meta had announced that it was shutting off quest VR access to Horizon Worlds, which doesn't, you know, in the context of this is kind of silly.

00:33:06:10 - 00:33:26:03
Unknown
It feels insignificant, actually. They walked it back. Bosworth yesterday. Bosworth yesterday basically said, oh, actually, no, we're not going to do it for the time being. So maybe they got enough push back. But but I think it's significant in the sense that they had a real strong idea of what they wanted the company priorities to be. They went down that direction.

00:33:26:09 - 00:33:43:29
Unknown
They're now they change the name of the company, they changed the name of the company to it. And now they're shifting directions and they're they've got their massive hiring spree with superintelligence, yet they're laying off 20% over here and that direction. And Alexander Wang is going to be in charge of everything. Now, there's been a reorganization. So he's only in charge of this group.

00:33:43:29 - 00:34:05:04
Unknown
And that group alone is losing 600 people. And he's trying to put a happy picture on that. And people playing about him as a bad young manager. The whole meta strategy metaverse strategy is obviously called, at the minimum, if not dying. You know, what are you gonna do to change the name of the company next to malt?

00:34:05:06 - 00:34:33:29
Unknown
Right. They buy more book, which is which? They're buying nothing. So they're going to change the name of the company to claw. Yeah, it wasn't going to. Yeah, but it's the current. Yeah. Name that and stop on the bandwagon. And they delayed because they delayed their models. Clearly is an admission that they're not there. So, yeah, it feels like, is this the new AOL?

00:34:34:01 - 00:35:04:05
Unknown
Yeah. Yeah, I don't know. It is. It is very interesting. They also actually along these lines have the avocado model, which was their next flagship model that they were supposedly working on and had plans to release, possibly even by now. Apparently that's delaying to May, which isn't too far from now, but they're basically saying like it's not up to snuff, it's not up to par with some of the others that are, you know, all the others that are out right now.

00:35:04:05 - 00:35:32:04
Unknown
They're all playing a game of what is what is the game where they all want to, you know, stay on top. Is it? I don't know, it's like they won't release their model until they know for sure that they have that two weeks where they can say, oh, see, we're topping the benchmarks before they get ousted. But I think what's interesting about that story is that there's even talk that there are thinking about or considering licensing Gemini inside of meta to stay competitive.

00:35:32:06 - 00:35:55:12
Unknown
In the meantime, Apple and Meta both licensed Gemini. That's that's fascinating. I mean, that's a that's a bit of a flashpoint that happens. Yeah. That's fascinating. Well, I think there's other there's other paradigm shifts going on. Right. And we'll get to this in a second. With open AI, everybody has anthropic jealousy. Yes. Because there's a lot in the strategy.

00:35:55:12 - 00:36:25:13
Unknown
Right. And then we also have the scale versus world models, balance going on. A paper that I read this week that was fascinating by four authors, including Yann LeCun, challenged this notion of AGI. Yay! I always get happy when that happens. Somebody challenges in arguing that human intelligence is the wrong goal, and general intelligence is B.S., that even people don't have a general one, even though, yes, we can learn things, we don't have general intelligence, we can go to.

00:36:25:18 - 00:36:48:09
Unknown
Every one of us is good at some stuff and crappy another. Yes. Oh, that's so true. Yeah. Right. Absolutely. And we're all crappy at chess compared to the machine now. And, and he has this wonderful argument in there that if you had one model try to do protein folding, but also required it to do everything and also put in folding laundry, it's clearly going to be less than a model that only does protein folding.

00:36:48:11 - 00:37:09:00
Unknown
And so besides role models, he's arguing for specialized models. He's arguing for not being this general do everything huge gigantic model. And and he's clearly he just has a new version of his, of his GPA, out, where he's trying to get to this idea that it doesn't predict pixels. It doesn't text, it tries to understand concepts in the video.

00:37:09:02 - 00:37:31:17
Unknown
You know, that's a chair is what a chair does. Right? So my point is that's a paradigm shift. And, you know, even DeepMind even there was a hot sauce. I'll see, because I was, I was thank you. DeepMind's but all in the scale thing. And they started talking world models. Yann is obviously fully on your world models.

00:37:31:24 - 00:37:54:06
Unknown
Feifei Lee raised a lot of money for role models, so that's happening. So what's kind of happened to, meta and Microsoft and OpenAI is we'll get to in a second. In that role where they're, they're so committed to one way. So thinking and then, oops, there's some stuff happening over here. And, it's going to be fascinating to watch.

00:37:54:06 - 00:38:16:09
Unknown
Who keeps up? Yes. Yes, indeed. Yeah. I'll be curious to see if this whole Gemini thing happens. I feel like that's really interesting, really telling development if that were to happen and, and I, you know, I know that the, the news about this was that it was would be a temporarily a temporary use. But I wonder if that would end up being the case.

00:38:16:11 - 00:38:42:23
Unknown
Well, it's it's not unlike where was it here. I had so the Wall Street Journal did a story. Apple's cheap, I bet could pay off big. What the iPhone maker has that others like is on parallel customer base. Okay, we know that that's the same as Google too. So what they're arguing basically is instead of investing all this money, it says that as Apple, fumbled the AI revolution.

00:38:42:29 - 00:39:07:29
Unknown
Not at all. It is making a cold. I'd bet that the most frenzied build out in the history of American capitalism will produce inadequate results. So they're arguing here, and that there's a wisdom to holding back. Now, meta has already invested too much. They invested? Yeah, $800 billion, whatever it is. And and others, their chips in the center, they're pushing these huge data centers and all of that stuff.

00:39:08:02 - 00:39:34:21
Unknown
So they're out there. Their money is already spent, on these bets, but Apple isn't. And in either case, using Gemini is not a wise. You give your customers what they want for the best of what you have now in a game of leapfrog. Okay, totally. But if it ends up being a big thing, will you wish you had gone all that's the question before I then that really is the other side.

00:39:34:22 - 00:39:57:00
Unknown
You also got to ask where your where your financial strength is. And in the case of Apple it is it is your journals right. It's with their customer base and the lock in their. In the case of Google, I would argue the same thing though they are investing. I don't think meta has strong lock in. But it doesn't have that overall view of the customer that it tried to have no everything platform.

00:39:57:02 - 00:40:17:26
Unknown
Microsoft one could argue does have some of that, but it's, it's, it's their own audience. I mean I'm not a Microsoft user so I don't, I don't like people who are. And I know a lot of people who are I certainly yeah. Even on this, this trip I met, quite, you know, at least a couple of people who are, who are hardcore as far as that's concerned.

00:40:18:00 - 00:40:42:00
Unknown
You look at their, at their, windows machine with empathy. Is that what it was? No. I just notice, you know, everybody has their, their, their preferred method and their preferred approach. And, copilot is never on my radar because I don't have, windows or, you know, Microsoft machines, but it was, it was interesting to just be with someone who did that was that was her approach.

00:40:42:00 - 00:41:09:05
Unknown
That was her style. And, you know, kind of learn a few things or two. When I think of meta, I, I don't think of the affinity aspect. No, I don't yeah, I agree, but that's because of people's affinity for others. I mean, Microsoft I mean yeah. Go ahead. No, I just like when I think about like what I would use all of them for the only reason I ever end up using Facebook or Meta AI is because I happen to be wearing their smart glasses.

00:41:09:07 - 00:41:37:24
Unknown
You know, like I would never it never occurred to me to open up meta AI and use it for a particular thing. And I also don't hear people talking about that, you know, it's just not part of that conversation. So I wonder how much that's actually happening. Before Gemini, I got as broadly offered as it is now. I for about, I would say listening to more than three weeks I would regularly open but I it was just easy metadata I boom I asked my question, I get it, I ask for a definition, I ask for something else.

00:41:37:29 - 00:42:00:23
Unknown
I got the same basic services I could get from any of them, and I tend to use it oddly more than, open AI. Yeah, okay. But that went away. Yeah, yeah. And I don't see any integration of the integration that there is of AI in the Facebook sucks because it's, it's I made pictures and videos trying to fool you.

00:42:01:00 - 00:42:23:04
Unknown
It's crap, right. Meta. Did you know acquire mantis. So they've they've got that going for them. You know maybe that ends up being a part of it. No China's not happy about that. So who knows what happens next there. Yeah who knows what happens there. That's a good point. I think mantis even, released. We don't have it in the rundown, but like, on my computer.

00:42:23:06 - 00:42:45:13
Unknown
Yeah. Everybody's now doing their their, tendrils into the computer, you know, for the on device stuff. And mantis is doing that as well, which means meta is doing that. Because meta does officially own mantis now, so. Yeah, I think so. Yeah. Well, then you've got OpenAI, which as you, as you alluded to, OpenAI, shifting.

00:42:45:16 - 00:43:10:22
Unknown
Well, there's few things happening with OpenAI right now, that we can talk about. First of all, there's Fujiko, who calls side quests, calls out side quests in OpenAI strategy and basically says, like, hey, we need to focus on our core business and not do all of these side quests. I, I don't know what that actually refers to because OpenAI is constantly releasing things.

00:43:10:22 - 00:43:43:27
Unknown
So what what are the non side quests? Is that coding? Is that their their standard level approach is that Jonny I've is it is it the hardware? I don't know I don't know I kind of feels like a side quest potentially. It does I think stage it certainly does. I interpret this as anthropic entry. Yeah I think this is think coding and B2B even even as open dialog takes us to be the see in a whole new world that operates so quickly, where OpenAI could have been in a better position in a way, because it has a bigger consumer presence of mind.

00:43:43:29 - 00:44:13:14
Unknown
But now they're going to push into coding and enterprises, because that's where anthropic is succeeding. Well, and they also, just made an acquisition, or announced an acquisition along that line as well. Sorry. Let's see here. Nope. That's wrong. There we go. Astral OpenAI acquiring, astral, which is, which is a service that's all about Python development and folding that into Codex.

00:44:13:16 - 00:44:36:16
Unknown
And I think this is exactly what you're talking about. Right. Like there's there is a lot of, anthropic envy in, in the realm of or in the AI world right now, because anthropic is getting the, the enterprise deals, anthropic coding kind of integrations and infrastructure is really impressive and accessible, very accessible. And, I absolutely I think OpenAI wants a piece of that.

00:44:36:16 - 00:45:09:09
Unknown
So this astral, acquisition certainly brings them a bit closer to anthropic and even cursor, which I you know, if you're watching the video version, they kind of jumped, jumped the gun on the cursor news, which, they also, announced an updated composer to coding model. So a lot of code focus right now, which I think I think you're absolutely right, which is very feels very reactive to the fact that I'm getting a lot of this right, getting a lot of these deals, and they all want a piece of the pie.

00:45:09:11 - 00:45:37:03
Unknown
Yep. Yeah, yeah. So, okay. Do we think that, ChatGPT is going to be able to or OpenAI is going to be able to pull that off with ChatGPT? I think that the there's the level of the desperation Ometer I'm sure we should patent that desperation. Ometer. So 0 to 10, right? Yeah, Nvidia is a one because people are starting to say, oh, what are they going to do about inference while they solve that?

00:45:37:03 - 00:45:56:15
Unknown
So but they responded. So there was a little bit of desperation there. But hand it. Yeah. The Rock purchase, the grok deal. Over on the other end. I wouldn't put meta out of ten, but I might put them in at nine. Yeah. I mean, they're up there. Yeah. All right. So in the desperation ometer, where would you put up an AI?

00:45:56:17 - 00:46:23:27
Unknown
Desperation. Yeah. They got to be somewhere in the middle. Although I don't know the I don't know that they would admit that there. Oh, I don't know where to beat at all. I know they're higher than I think they're, 7 or 8. 7 or 8. Wow. Okay. That high. Yeah, yeah, I think I think I mean there is, you know, it often comes back around to kind of the moat, the moat conversation, which is that Google has a solid moat.

00:46:23:27 - 00:46:48:26
Unknown
You know, some of these other competitors have plenty of other things to fund themselves through. Meanwhile, you've got OpenAI really driving its business by acquisitions, by announcing new things. And maybe this is what Fugees is talking about. But, what do they have to fall back on that isn't this has to work? You know, the other problem is, as I as I can, I don't think anything has happened with their Stargate project.

00:46:48:28 - 00:47:13:00
Unknown
Right. We're having a bit of a $500 billion effort. It was going to include Oracle which has its own desperation and trouble. SoftBank which is its own trouble because of where they're investing. And so OpenAI has to go out and basically buy on the open market with it rather than owning it as much. They've got problems, I think, because they're so in the scale.

00:47:13:00 - 00:47:32:15
Unknown
Scale is going to solve it all. Scale is going to get us there. They believe in AGI, blah, blah, blah, blah, blah. They're committed to that path. And I don't they're concerned economically. Good path. What do you think about their their adult mode. Do you think this is the path to the success. No.

00:47:32:22 - 00:47:56:14
Unknown
What do you think. Oh I don't know. I think this is yeah. I mean are we, are we actually going to see this roll out. Internally they've got safety experts that are saying like, hey, wait a minute. These things are actually can be somewhat difficult. Of course, or not difficult, but dangerous. Yes. They are citing a case where a user forms a bond with the chat bot.

00:47:56:16 - 00:48:15:12
Unknown
Apparently there was a case where that ultimately led someone to take their own life. I think this ties in with what you and I have talked about plenty of times before, which is it's not always necessarily the direct fault of the the bot itself, but maybe the kind of the preexisting setup of the user who, who is using it.

00:48:15:12 - 00:48:43:25
Unknown
So how many times is this happening? I don't know, but, do I think that OpenAI leveraging an adult mode is like a savior for its company? No, I don't I mean, has it has it been a boon for for grok, for AI? I know it's gotten headlines, but I don't know no one in my life, you know, like in my, in my life is, is, is talking about how that's the one that they use because of that feature.

00:48:43:25 - 00:49:10:12
Unknown
But maybe you just don't talk about that feature. That's. Well, that's true. I mean, I got to go back to my, my, my history lessons here at Gutenberg. Watch out. Yeah. Is that is that when print came along, and it wasn't just scribes, silent reading came along and silent writing came along, and thus heretical writing and pornography came along.

00:49:10:15 - 00:49:33:08
Unknown
If you had to dictate your thoughts to a scribe, you were unlikely to dictate porn or unlikely to dictate that which will get you burned at the stake, either might. But once you could produce and consume silently, that led to poor, people are going to use this anyway for the stuff I just don't see open AI making this a business option.

00:49:33:10 - 00:49:57:14
Unknown
It's something that no, I don't think Elon Musk and Grok would do just to be provocative. Let him. Yes. Yeah. Well said. Totally I yeah. You put words to kind of what I was having a hard time articulating an AI generated version of Val Kilmer. So Val Kilmer, movie star, famous movie star who passed away, very recently got wounded.

00:49:57:14 - 00:50:24:14
Unknown
He passed away last year. Maybe even, he was in line to April of last year. Okay. Yeah. So almost, almost exactly a year ago, he was set to star in an indie film called As Deep as the grave. Man, that the the tie in of the name of the film with the topic of this story is, is is eerily, you know, coincidental.

00:50:24:14 - 00:50:45:27
Unknown
But they had him lined up to play a role that Val Kilmer was apparently very passionate about. It tied into his background, his, I think he has Native American heritage. So it was kind of a tie in to that. But he passed away before he could star in the film, for the filmmakers had shot all of this footage.

00:50:46:00 - 00:51:22:03
Unknown
It basically shot the film. And what they didn't have was his role because he was not there to act in it. And they don't they didn't have the money to hire someone new and roll new film to fill in all those pieces. Nor do they want to, because Kilmer wanted to be involved in this. So they went to Kilmer's estate, Kilmer's estate, which is, you know, his family, his son and daughter and a few others knew that Val Kilmer had a passion for this project, so they ended up ultimately greenlighting it and supplied images and video from throughout Val Kilmer his life.

00:51:22:06 - 00:51:42:12
Unknown
For generative AI to train on. And, you know, they're going to even include his damaged voice. So what did he what did he die from? He said he got he got pneumonia, which I didn't know. Was that kind of a sudden thing. People knew there was some sort of. Yeah, condition tied to that, I think, but I can't recall off the top of my head.

00:51:42:12 - 00:51:59:29
Unknown
But anyways, he had a his voice was very, very damaged from, from the health issues he had over the last couple of years. So they were going to integrate that into, into the generative throat cancer. I'm sorry. Right. Throat cancer. Throat cancer okay. There we go. There we go. And so anyways they got the green light. And so Kilmer is going to be in a number of scenes.

00:51:59:29 - 00:52:22:17
Unknown
This isn't just like, Carrie Fisher in a scene in the rise of of of Skywalker, you know, Star Wars, after she passed away. This is like he's he's like a supporting character throughout. So this will be a film with Val Kilmer, via generative AI. We're really curious to see how this goes down. So what do you think?

00:52:22:19 - 00:52:41:06
Unknown
I don't know, I mean, I mean, on one hand, like, I'm not immediately, you know, they checked in with with their, with the estate. The estate gave the thumbs, thumbs up. Does that, you know, line up with what Val Kilmer would have wanted. I mean, by all accounts, it sounds, you know, at least reading through this article.

00:52:41:06 - 00:53:06:20
Unknown
Anyways, that, Kilmer was really, really, all in on this. So the estate does a bit for you when you're gone and the estate speaks for you. Deputize them. Yeah. So will the quality be able to say no? Yeah. Right. But I can imagine. But it puts actors in a really odd spot because they all want to object to using AI to replace them.

00:53:06:22 - 00:53:23:07
Unknown
But here is, oh, it's Val Kilmer is dying. Wish you really want to be against that, Yeah. Right. So it creates a fascinating precedent in this debate. Does, I mean, you have to die to be recreated in the AI, or can you just make up a new character in AI? Obviously, there's all kinds of variations there.

00:53:23:09 - 00:53:39:22
Unknown
And we're going to see all those things. We're obviously going to see it, see it all play out. Yeah, it's going to start with extras. It's going to start with commercial actors. Yeah. But that this is a pretty much it's going to look like that. That's what I mean. Like it's a major role with a very recognizable actor.

00:53:39:24 - 00:54:00:09
Unknown
So if there's if there's a reason for people to look and find the tells that make them hate that this happened because, oh my God, it looks awful. Or whatever the case may be. This has all the all the signs that would point them in that direction. Right? Like they're going to it's recognized it's a recognizable person. So you're going to know whether it feels on or off.

00:54:00:10 - 00:54:19:25
Unknown
You're going to be in there. You got a suspension of disbelief. Is that is the heart of all movies? Yeah. Right. Yeah. So you're sitting in that movie theater and, you know, this is the case. You know, it's not Val Kilmer. I think as a as a moviegoer, you're going to be constantly judging and looking for those tell.

00:54:19:25 - 00:54:38:25
Unknown
Yeah, I mean, that's that's what I think. You're absolutely right. Going into a film like this, I'd be less focused on the story when he's on the, on the screen and more focused on the what, you know, looking for the tells or, and does it feel on or does it feel off or whatever? Yeah, I bet, but I'm curious.

00:54:38:28 - 00:55:08:22
Unknown
I'm definitely curious to see how it all plays out. I don't get the epic factor from the estate, you know, permission perspective. Although no one truly knows whether he would have been okay with this, this specifically. But, I kind of feel like they did their due diligence there, and it feels okay there. You know, what is the difference between Val Kilmer as an actor, a virtual actor in a film versus the millions of CGI characters that we've seen before?

00:55:08:24 - 00:55:29:18
Unknown
Right. You know, they just weren't meant to be humans that, you know, were highly, high quality reproductions of a human being. That's that, I guess, is the difference. Will we be at a point and are we at that point to where that, that uncanny valley is no longer a thing? You know, I don't know.

00:55:29:19 - 00:55:50:10
Unknown
You're very welcome. Be, and look at the earlier story about end video and their new their new, yes. Whatever they call it, the brand. Totally the GPU. Yeah. Thank you. We also have the case of Michael Jackson at the Billboard Music Awards as a hologram. Yeah. What year was this? That was 2000, 14,000 years ago.

00:55:50:10 - 00:56:08:21
Unknown
Yeah, right. But that that's a little different because that, you know, it's a hologram. It looks like a hologram. And it's meant to be a tribute to him. And you don't have to rest a plot and character on top of it. It's just, let's remember Michael Jackson. What a cool way to do that. Okay. Done him right. And he was on a bit.

00:56:08:23 - 00:56:24:21
Unknown
The large stage was easy to kind of like put him on this scale where he's, you know, this, I mean, this is, you know, this is a screenshot from the film. This is like the first release, generative AI reproduction of Val Kilmer in this film. And so you can imagine these are going to be a lot of close up shots.

00:56:24:23 - 00:56:41:07
Unknown
Then let's just say there's going to be a ton of scrutiny on this. Oh yeah. And there's going to be a lot of people who refused to watch it specifically because it has anything to do with generative AI. And yeah, yeah, it'll be an interesting case to follow. That's, I think, what captured my, my attention on that one.

00:56:41:09 - 00:57:10:09
Unknown
And then we got an encyclopedia. There's not a whole lot to say here except, Encyclopedia Britannica and Merriam-Webster. I can't show it because it's on, Reuters and I don't have a subscription. They're accusing OpenAI of copying 100,000 entries in their GPT training. Yeah, I get it cannibalized. Britannica web traffic. Yes. Fabricating site citations, offering, incorrect citations.

00:57:10:11 - 00:57:39:23
Unknown
OpenAI. I said our models in power innovation are trained on publicly available data and grounded in fair use. Yeah, I can imagine that. Unlike the anthropic bookcase, OpenAI, I can find one person inside the company who owns a subscription to Britannica Online and to Webster's Online. Right. And then and then that might do it in. The other thing is, and I have great respect for Webster's, Britannica is a bit, is outmoded.

00:57:39:26 - 00:58:00:26
Unknown
I was at an event, many years that was so that was a lot of the comments I saw, by the way, which is like, oh my goodness, they're still around. I still run, right? I was at an event many years ago, at Harvard, where the then deputy editor of the New York Times, Jill Abramson, briefly editor, and, Jimmy Wales were and and Abramson was giving the you don't understand.

00:58:00:26 - 00:58:27:01
Unknown
This is during the Iraq War. You don't understand how expensive it is to have a Baghdad bureau and know what the value of journalism, blah, blah, blah. And Jimmy, it is inimitable. Fashion spoke up and said, nobody cares. Nobody thought that a whole bunch of amateurs could replace Encyclopedia Britannica. And they did. And so there's more value at, Wikipedia than there is.

00:58:27:01 - 00:58:55:11
Unknown
It is likely to be redundant right now. It's, it's it's outdated. It's small. It's it's limited. It's not as diverse, as a as another source. Fine. It has value. Right. So I'm not saying it doesn't, does. But there are other sources of information. It's not necessary dictionary a lot of dictionaries. But there again, if somebody owns a copy of the Webster's Dictionary, under the anthropic decision, as long as they owned it, that legitimately and acquired it legitimately.

00:58:55:14 - 00:59:14:12
Unknown
Cool. So, I it's the same as the, the we haven't really talked about the Grammarly case where Grammarly, yes. Oh, yeah. We didn't, because it kind of felt like we did. I know I did on a show, I can't remember if it was on the show or this one, but the brief of it is that Grammarly, without permission?

00:59:14:14 - 00:59:32:05
Unknown
Yes. We just very we did. Okay. So that's authors and and now you have, one of the authors is making a big deal that she's filing a class action suit, and this is evil and awful. I think I do say this on the show, if Grammarly had just said, well, you should read Jay Rosen about this because he wrote something relevant and linked to it.

00:59:32:05 - 00:59:56:14
Unknown
It'd be like a cert that'd be a smart search engine. They could have done it smartly and they didn't. And they pissed off, entitled self-important writers. Okay. They did it stupidly, I agree. Is that worth the whole suit? Is that worth a revolution? I don't think so. The value of content as content I wrote about this in the Gutenberg parenthesis is diminished constantly.

00:59:56:14 - 01:00:19:22
Unknown
It was diminished before AI diminishes the more. That's not where the value lies. It's not a it's not a, the idea of copyright was started by the industry of booksellers and publishers to create a marketplace for, for content, for text as a tradable asset. The whole metaphor of it is property is an invention, and it doesn't have to be the way we look at it forever.

01:00:19:25 - 01:00:27:14
Unknown
So, yeah, I think, Britannica used to love you, babe, but,

01:00:27:17 - 01:00:52:19
Unknown
Yeah, yeah it is. Fortunately, it seems like time has kind of moved on. Yep. Yeah, yeah, I, I agree with that one. We have a YouTube channel. You go to youtube.com and, search for AI inside or AI inside show. You will find our channel and you might want to do that because if you like listening to this podcast, well, you can get it with video and links and all that kind of stuff shown on the screen as well.

01:00:52:19 - 01:01:07:21
Unknown
So check it out. Also, our interviews are a heck of a lot of fun to watch. We know you enjoy listening to them, but you can watch them. We learn something. Every one of them. Indeed we do. I'm going to take a quick break. Then we got a little speed round a bunch of speed round items that we can blast through real quick.

01:01:07:21 - 01:01:10:27
Unknown
That's coming up here in a moment.

01:01:10:29 - 01:01:38:18
Unknown
All right, Google, little trifecta here. Google is rolling out personal intelligence features. To all paid. Yes it was. It was for paid users, but now also all free users. So the personal intelligence features basically take, aspects of your Google account, like your Gmail, your YouTube, your you know, other things, your Google Drive and all that stuff and kind of integrates it into the Gmail or the Gemini experience.

01:01:38:21 - 01:01:59:01
Unknown
It is opt in. So and it's kind of granularly opt in, last I checked, so that you can say yes to this, no to that. I'm fine with maps being shared. But I don't want, you know, my drive to be shared. What? However you want to, focus that. But, free users will be able to get that important context, looped in if you choose.

01:01:59:04 - 01:02:19:22
Unknown
I find it to be useful sometimes, you know, not all the times, but sometimes where I want to use it. Of course I'm going a wine. I can't use it, which is in Chrome. Yeah, but I did that. I did that one in last week. I'll spare you this week. I know week. Maybe you'll be able to, ask maps in, Google Maps update.

01:02:19:25 - 01:02:44:09
Unknown
Because that's on your smartphone. I'm assuming that no issues there. Makes it more conversational so you can handle more complex, highly specific questions. Tied in also to some of your data, like using your reviews, your photos, your past activity. So you might find an updated, version of that on your phone at least Jeff Lee, such that I use Waze.

01:02:44:09 - 01:03:04:21
Unknown
I've got this. Now, I got to remember to use maps to try this out. Which do you use? I use maps, yeah, 100%. That's that's the app that I use. I haven't tested this out, this out specifically, but, Yeah, that's my navigation app. I tried Waze and I always, I always end up going, why, why, why am I using Waze?

01:03:04:25 - 01:03:19:24
Unknown
Like what? What is the reason versus just using other ways from the beginning? It's just I just feel terrible for me. So I think it's an affinity thing, like you're just used to what you are people and it's gotten better. I want to ask it for a Starbucks on my way. Used to be neither could really do that.

01:03:19:24 - 01:03:36:10
Unknown
Now they're smarter questions like that. They're smarter ways might be better than maps. Sometimes maps end up not doing a great job, but each good scroll one will say, oh, this is three miles away. And then it gives me a 30 mile route. Yes, it was three miles away. Or you should go there. It's three miles away. Oh, wait a minute.

01:03:36:10 - 01:03:58:01
Unknown
It's back there. Three miles. I already passed it, you know, instead of knowing smartly, like I'm not going to turn around, just give me the next one that's in front of me. Like that's a switch I need. Let's see here. And then Google also launched just today an update to its stitch tool. And so this is basically that, you know they've they've highlighted here vibe design.

01:03:58:01 - 01:04:23:14
Unknown
So this is all about UI design. And stitch isn't you know this is not a new design tool. But what they've done is they've taken as they're doing with everything bringing their, their AI, native, you know, kind of language, conversational approach to UI design. So you can, you know, design on a canvas, build, software design prototypes, that sort of thing.

01:04:23:16 - 01:04:46:02
Unknown
You can vibe design as a, as it were. So I think developers will, find that interesting. OpenAI not a lot to say about this with a release GPT 5.4, Mini and Nano. So yes, you can imagine they are smaller. They are faster. They're really tuned for, you know, they're not meant to be best at everything.

01:04:46:02 - 01:05:18:10
Unknown
They're just meant to be ridiculously fast and inexpensive for developers to tap into. So duly noted on the record. There you go. There you go. Always something new. Let's see here a U.S. appeals court has temporarily frozen a lower court order that blocked perplexity from running its shopping agents on logged in Amazon accounts. So perplexity is, of course, pushing for a longer term pause throughout the appeal of the lower court's decision.

01:05:18:12 - 01:05:37:05
Unknown
So there's that. I think for the time being that I guess that means that you can use your perplexity agent to, put things into your, Amazon Prime shopping cart until the next twist and turn happens. I was I will make money from that. What's not to love, I don't hear yeah, I don't get it like that.

01:05:37:11 - 01:06:04:17
Unknown
Because they want you to use their thing because I don't know. So that's probably probably what that's all about. The AI darling. Anthropic. For now. Auto generate. Yeah. This week, this time around, auto generates interactive chats, diagrams and other visuals now inside of the chat so it detects when those methods actually make sense. Although as a user, you can request for those things to, to appear inside of your chat.

01:06:04:17 - 01:06:26:04
Unknown
You can edit them on the fly. Kind of makes those charts and those visual elements a bit more, attached into the conversation or transient, as compared to something like artifacts that are a bit more permanent. So that's one thing I really love about cloud. I love how visual it is. I'm a visual learner and it does a really great job.

01:06:26:04 - 01:06:43:28
Unknown
When I say, put this into a visual, presentation or put this into a PDF, you know that looks really nice or whatever. It doesn't get it perfect. None of them do. But I think it does a really great job, and it really helps my brain to to learn some of the things that I'm using it for. You know, that's kind of miraculous.

01:06:43:28 - 01:07:12:06
Unknown
Still, it is pretty damned amazing. Yeah, it's pretty cool that that's all it takes. And then finally, Microsoft is launching Copilot Health Consumer Copilot feature. That is going to take in medical records of wearable data, deliver health advice. So they're framing this as a path to medical, quote, superintelligence, medical superintelligence. Now there's going to be all different types of superintelligence, initial US rollout with paid plans coming in the future.

01:07:12:09 - 01:07:34:03
Unknown
Would you trust a Microsoft Copilot health? That is the. Well, at the same time, we should mention very quickly that Google had a health, service that was giving you advice from other people, just plain old people on Reddit. They wisely killed it this week. Oh, I missed that one. Yeah. What do you do? No, no, no, the old joke that you want to do.

01:07:34:03 - 01:08:02:10
Unknown
You want a citizen surgeon? Do you want a surgeon? Yeah. No, no, no. Definitely not. That sounds like a horrible idea. You mentioned I just happened to be scrolling by my Facebook. Alan Rusbridger, former editor of The Guardian, puts up a picture from a charity shop. The full set 24 volumes of the 1920s edition of, the Encyclopedia Britannica and the store labels it perfect for interior design project $100 for the set.

01:08:02:13 - 01:08:28:27
Unknown
So they they've just become visual relics or AI company. You could buy it and then just, scan it and it's yours. And then you're soul. You're you're solid. Yeah. You've covered your bases. There you go. It only costs you 100 bucks. But now what's covering the television? Know what the hell? Yeah, yeah. So, Jeff, normally we would say go to Jeff jarvis.com for and we still do go there.

01:08:29:01 - 01:08:49:18
Unknown
You can find all of Jeff's books, but you've got, some book news to announce. I got something you want to know. So if you go to my medium page, you'll see that I'm announcing a new book series relevant to the show called intelligence, AI and humanity, working with Bloomsbury Academic, which is the publisher of my, three books, three of my books, Gutenberg and Magazine.

01:08:49:18 - 01:09:23:06
Unknown
And I type. And now, I'm humbled to say that I'm going to edit a new book series. The idea here is it's not about the technology of AI. It's about having writers from many disciplines come in and, reflect on how AI system reflects on society. And, we have the first three authors, signed up, I'm delighted to say one is, Rahman Chowdhury, who, was the head of, ethics at Twitter until she had the honor of being fired by, Elon Musk.

01:09:23:08 - 01:09:52:27
Unknown
She worked for the Senate. She she is an academic of, a political scientist. She's just an amazing work. She's going to write a book. Asking basically what is intelligence, which is the perfect beginning of the series. Because what AI does is make us ask those kinds of questions. Doctor Charlotte McElwain, who's the vice provost and professor at NYU, author of the wonderful book Black Software, is going to write a book surprisingly optimistic about, race and AI and how AI.

01:09:52:28 - 01:10:17:26
Unknown
This goes back to our conversation earlier about the democratization that comes along with Opening Claw. Might offer, Black America a chance to, find some retribution for the technological oppression that's occurred before. And then Doctor Matthew Kirschenbaum, now at the University of Virginia, formerly University of Maryland, wrote an essay that I've, linked to before here called Text Apocalypse on the Atlantic.

01:10:17:26 - 01:10:36:25
Unknown
He's turning that into a book. Yes. About our relationship to text. So those are the first three books that all have been that is published and won't come out till next year, but, early next year and then, I'm going to be looking for more ideas and more authors to come to me for things like wondering, what does I tell us about our views of learning?

01:10:36:28 - 01:10:57:03
Unknown
What does it tell us about our sense of creativity? What does it tell us about, God? And, thinking that we can create life? So I want to look at a lot of opportunities to talk to interested authors from around the world, diverse authors who with diverse perspectives and from many disciplines. And that's going to be fun.

01:10:57:03 - 01:11:26:22
Unknown
I am humbled, you know, here is an academic publisher and I'm a fake academic, but, by my editor and publisher, Horace Naqvi, called one day and said, I think I want to start a book series, but hey, how you want to edit it? And I mean, amazing. Yes. So heck yeah. So, knowing little I'm on to I podcast that I have a book series on so I'm just living the I this is how you know, this is how you continue to learn and grow and and everything.

01:11:26:22 - 01:11:43:03
Unknown
I think I think you're the perfect person for this. I can't wait to see it. That's awesome. That's awesome. So if people want to follow or find this detail, they go to medium and search Jeff Jarvis or how do you find the medium? And then there's a link there to the, Bloomsbury page. And just watch the space.

01:11:43:03 - 01:11:59:29
Unknown
I'll, you know, you know, I mean, I'll be plugging away. Always be so. Yeah. Of course. So we're going to do some events later which might involve Jason. Okay. We're talking about that. And, so just watch the space and we'll tell you all about it. Love it, love it. Everybody stay tuned for that. Excellent stuff. Congratulations.

01:12:00:00 - 01:12:21:04
Unknown
Thank you. Yet another, project I love it. Yep. You can go to AI inside Dot show for all of the links to everything that we do for this show. All the ways to subscribe, reviews, fact. We do have even more reviews that have appeared. That's amazing. Yeah. Thank you folks. You busy? I didn't even know that two more came through.

01:12:21:07 - 01:12:44:22
Unknown
Jason and Jeff cover relevant and timely topics each week with clear and thoughtful discourse. Thank you, thank you. Can't read the next gray on BCP 2600. Hans, zoom in. We've got Jeffrey married. Cheyney. Yeah. Your name there? Tom. Roughly. And the Apple loyalist. And I think that is let's see here. Oh that's a that's that one's a little bit older so yeah.

01:12:44:22 - 01:13:05:28
Unknown
So Tom. Yes. Thank you folks. Tom Jeffrey Hans and BCP 2600. Thank you so much for the reviews. That really really helps out. And I appreciate that that you did that. So thank you. Patreon.com slash AI inside show if you want to support us on a deeper level, that's how you can do it there. We do have some amazing executive producers.

01:13:05:28 - 01:13:30:07
Unknown
Doctor do, Jeffrey, Mario, Jeannie previously mentioned, radio Asheville, one of 3.7 Dante Saint James bond. Bonner. Derek. Jason knife for Jason Brady, Anthony Downs, Mark Archer and Carsten Szymanski. Some amazing, patrons who, you know, all of you are amazing. These folk go, just above and beyond. So thank you so much for your continued support. And finally, thank you so much.

01:13:30:10 - 01:13:52:01
Unknown
To Victor Bogue, not and Daniel Croft for video support behind the scenes. It just really, really helps helps me out is my beauty diehards, voice sometimes too busy. But I appreciate, Victor and Daniel deeply so. And I appreciate you, Jeff. Thank you. Learn a lot doing the show with you. And, Yeah, we'll see you all next time on another episode of inside.

01:13:52:08 - 01:13:53:12
Unknown
Take care everybody, by.