Rope-a-Dope

SPEAKER_00 0:04

The Rumble in the Jungle. October 30th, 1974, Kinshasa Zaire. George Foreman, the most feared man on the planet, had destroyed Fraser and Norton, both in two rounds, and nobody gave Mohammed Ali a chance. Ali's plan looked from the outside like losing. Go to the ropes, cover up. Let Foreman throw everything he had until his arms stopped working. Then in the eighth, come off the ropes and put him on the canvas. The rope adopt. Absorb the punishment, let your opponent exhaust themselves, and then finish it. So is that what is playing out between the two leading AI Frontier labs just now? Which one's on the ropes? Who's going to be left standing? Welcome to the AI transition. I'm Steven, and always Lauren.

SPEAKER_01 0:47

Oh my god, Steven, that was big. Is this float like a butterfly, sting like a Chat GPT?

SPEAKER_00 0:52

Oh, I like that one. I like that one.

SPEAKER_01 0:56

Boxing, we've gone into boxing, Steve. Unexpected.

SPEAKER_00 1:00

Well, it is the analogy that kind of came up for me this week when you're looking at these two BMOS that are kind of going at it these days. Um, although I I must admit, um my my wife did say that I'd be showing my age talking about you know a boxing match from 1974. But hey, hopefully it's in the zeitgeist out there somewhere.

SPEAKER_01 1:18

One of us was born then, Stephen. Let's not go there. We walked into that one. Uh Burn, I've got a good pun for that one later on. You know, George Foreman is the is the clue there. So who's on the ropes, Stephen? Who's who's uh where's this where's this leading us to? What are we chatting about today?

SPEAKER_00 1:38

Well, we're talking about open AI and anthropic. So the makers of ChatGPT and Claude. Now, there are more um AI LLMs that are out there, but these are the two that are really in the news just now and are ones that are really making the big waves as well. What's happening here is that hype is beginning to meet reality. So the first part we're going to talk about is what has actually changed in the last couple of months. And we we can talk about you know what's changed for us because I think it's significant in the last couple of months. You know, we've seen this on the ground. But also let's talk about what the big picture, because I think there's some real big fundamentals that have moved.

SPEAKER_01 2:18

Uh, there's been some huge things in the last few weeks where you've got, I'm trying to, I can't remember the box, the analogies that we've lined up here, but you know, one of them has actually lost functionality, the other one's just come in with some really huge leaps. So there's so many huge things going on at the moment. We've even got some IPOs coming out as well, which we can, you know, dig into shortly. But some of the hype, like you said, it's meeting reality.

SPEAKER_00 2:40

Uh so let's dive into some of those specifics. So, in particular, about anthropic. So, anthropic are the ones that we've talked about, you know, the last couple of episodes and had the big beef with what was going on with the Department of War over in the USA.

SPEAKER_01 2:54

Well, this is a good stop it.

SPEAKER_00 2:56

Yeah, yeah, we did that one.

SPEAKER_01 2:59

Stopping myself. Anything I can sing, I'll do that. Yes, Department of War.

SPEAKER_00 3:04

So, but in the last four months, Anthropic have moved from an annualized revenue run rate, you know, how much money they're bringing in in the year, up from about 9 billion a year to over 30 billion a year. Now, that's quite a lot in four months, no matter what company it is. That's really significant.

SPEAKER_01 3:23

Massive, tripling in about four months. And there's lots going on to drive that, we think, or we're interested to kind of you know chat about it because you know we talked briefly, I think last time a little bit about the different models they had in place in terms of their revenue, and I think this is really coming home to roost.

SPEAKER_00 3:39

That's that's the really big thing because you've got over a thousand enterprise clients now with Anthropic, and they're paying more than a million dollars a year. And uh and I I think Anthropic are getting something like 80% of the revenues from business customers rather than consumers. So whereas you know, the opposite kind of what was going on with OpenAI, or they're about to make a pivot, which we'll come to later as well.

SPEAKER_01 4:03

Exactly. You really had OpenAI investing in a lot of those more consumer-based features, and you're suddenly seeing them drop and realizing, hang on a second, maybe the model's not quite landing in terms of that revenue.

SPEAKER_00 4:15

Yeah. The main driver behind of this is Claude Code, which is about software development. And look, we are going to talk about that today, but but for those people who are listening who aren't in the kind of the coding area, what this is showing is how it can expand and use that to move into so many other areas as well. So it's kind of like a a proof of concept in this space and a foundation that's in there that you're with this Claude Code, then you can move into legal product management, financial analysis, sales. It's all the same sort of stuff when you dig into it.

SPEAKER_01 4:48

That's huge. And just that that game changer of how interconnected you can feel with your other tools that you're used to, or having to go outside of this ecosystem and how all inclusive it is now. We're probably huge converts and we'll have to be careful of our buyers here, Stephen.

SPEAKER_00 5:03

Uh I know but look, I think that leads into that the you know what changed at the tail end of last year was you know, Anthropic started coming out with Cloud Code and then Cloud Cowork. And it it was moving away from that interface where we were for a number of years using Chat GPT like a fancy Google, if we're honest, right? Let's be honest. You're putting in the one-liner and you're getting, I was gonna say, you're getting the funny joke back. You're not getting the funny joke back.

SPEAKER_01 5:32

I am writing these jokes. They're the jokes coming out of AI are just rubbish. And I'm a fan of a rubbish, can't even get a groan out of them, they're so bad. Um but but yeah, just that integration, what it can do holistically, and even integration, that's like even you know, choose our words, it really does feel so different in terms of what you individually can do.

SPEAKER_00 5:50

Yeah, it's it's that it's that it's that real shift. And and look, uh for be transparent about you know, kind of my experience over the last couple of months, you know, heavily now into into to co-working Claude Code. And and it is a it is a game changer. It's a fundamentally different way than I was using AI in a year and even three months ago, to be honest. You know, it's building apps and using agents for actual work and automating large parts of you know, boring parts of work life and or even parts of personal life as well, that would have seemed like sci-fi not that long ago, but it actually does work. And you're you've been experiencing the same of it's like, oh, what's it?

SPEAKER_01 6:29

And I I think even that mindset, like when you're working with these tools, you're trying to take a step back and go, hang on, what can I actually get it to do for me versus which bit I should do myself? And once you're starting to learn these kind of paradigms and how it all connects, it's just so powerful in terms of I know it's really dorky, we all think of administrative tasks we can solve them, but once you start playing around with this and seeing what it can do there very quickly, you can get so much more creative.

SPEAKER_00 6:54

Yeah, and and probably what we're seeing now out in the consumer world, you know, that's already been happening a fair bit more the tail end of last year in the corporate world, which is probably why there's been this explosion in growth for anthropic being able to, as you said, you know, tripling in growth um just in the last four months. Because there's you know, there's a lot of organizations that are using. I think you you saw a McKinsey survey recently.

SPEAKER_01 7:20

Yeah, and look, it it's interesting too, because I think the numbers have remained kind of pretty stable. It's more how they're actually working with it. So there was a bit of a bit of a it was a global survey back in November last year where you've got about 90% of organizations using um AI in one business function. So we're often talking about these big corporates, right? Um however, you know, it it's a fraction up from about 80, but 10% is a pretty big leap in itself in about six months. Um, we've still only got about like 5% that you're calling these as AI high performers, where they're actually seeing, you know, meaningful profits or, you know, um uh optimizations. You're seeing this huge gap where you've got these organizations, those little five percenters, really embedding these tools and optimizing how they work in terms of those efficiencies and getting that vague creation. You've still got people kind of stuck in, I think they're calling it pilot purgatory, just to get some more puns, where they're still running some experiments but haven't kind of really figured out how to connect it. Um, but you're starting to see more and more actually moving. So I think there were some fresher stats from um Deloitte just last month where you've seen that jump up to about 30%, where you've got organizations meaningfully transforming. As in, we kind of get how to integrate this toolkit more. We're opening up our um our technology bases and integrating and seeing that, oh, hang on a second, we can really shorten our processes and and life cycle by implementing these tools.

SPEAKER_00 8:44

Yeah. And I think a lot of that was driven was maybe just over a year ago when Microsoft released co-pilot to everyone and basically gave everyone around the world co-pilot who worked in any sort of organization that used, you know, the the office suite. And uh people dabbled in it and played around with it, but it wasn't really useful in a kind of transformatory type sense. Whereas what you can do with the likes of um Joanthropic and Claude's, um, it can be transformatory of what's going on. So I think people are actually seeing some some real values beyond being able to create the latest image of the dog or cat um of their choice or the or the pretty picture. I think there's some actual useful stuff that's beginning to happen now, which is making this shift.

SPEAKER_01 9:33

Big time. And even I think you're starting to see some of these small and medium enterprises taking that leap because they're finding that there's, you know, ability to actually perform entirely new tasks. Often, like if you're working in a digital agency environment and you just want to play around with some product concepts, you're having to go out and engage some designers or some maybe some data experts or analysts, you're gonna, you know, build a bunch of you know, data to play around with and experiment. And now that's all within your hands, like that true analogy of four or five people that one person can actually go in there and experiment. So you're getting those nice little upticks and also the ability just to ramp up, you know, new starters faster, right? Like if you've got a skill gap, you know, generative AI is starting to really help. I think there's about 40% of these smaller organizations that said, look, it's actually starting to really help us with our skill gaps.

SPEAKER_00 10:21

That was a lot. It is a lot. I mean, I know we're throwing a lot of stats out there, but the overarching message there is that things are really shifting. They're they're fundamentally moving on at a at a rate where this is now becoming useful in a c uh in a corporate and a business entity now. Whereas probably, you know, a year ago it wasn't really, but a lot of that hype is now beginning to come true on the ground.

SPEAKER_01 10:43

Particularly where you're you're cutting that code and you're trying to deliver that value fast. What are some of the ways you've been um using it for Steven?

SPEAKER_00 10:51

So I you know originally was you know a developer, you know, been through the trenches with technology for a long part of my career before kind of drifted off into other things, you know, in in recent years. I I've I've I've found myself going back into that in in a in a biggish way, but not actually writing any code per se, but writing stuff that writes code, if that makes sense. So kind of that sort of abstraction layer above, and it works. So and this you know, we're gonna do some videos and some future podcasts on this l later about you know how you can actually do this yourselves. But basically being able to kind of be beyond just vibe coding, which is you know where you say, Hey, just create me an app, but there's actually structured ways of well, actually create me a specification that will go and create me an app, and then you've got both the documentation, the specification, and the app itself, and then you can build and iterate over that. And you can develop stuff that if you've got an idea but you don't know how to code it, you don't know the I'm I'm coding in a language which I've I don't know at all, right? And I'm integrating into systems where I have no idea what the APIs are, but I I know how to talk to it in order to be able to do it, and it works, and it works quickly as well.

SPEAKER_01 12:01

I think it's interesting too, kind of riding along a little bit with you as you're experimenting, where you've got that developer brain and you think, you know, particular kind of left to right with how you need to cut and test this code, and then you start to realise hang on a second, you've got um, you know, Claude Cowork telling you, actually, you don't need to do half of that. We can do all of this in here. And making that leap creatively is quite interesting.

SPEAKER_00 12:23

It's a whole different creative process which builds on top of a lot of the stuff that I've done before, but it is a different process as well, which is both uncomfortable and scary, right? Um, because there's there's big parts of it where I spent years learning this stuff in a previous life. Yeah. And and you don't need that now. Although it's good to know that conceptually, but being able to now use it just by these kind of conversations, um, yeah, it's it's really good stuff.

SPEAKER_01 12:55

And oh, am I that jerk now that who I'd never cut code, right? I grew up kind of as an analyst and always doing ways of working and delivery, and now it's like, well, we've evolved where I don't want to say never had to use it. See, these are critical thinking skills in terms of how you problem solve, but you can lean on some more of those design thinking, you know, concepts around, well, hang on, what could be something in this world we live in where we're all a little bit fearful and you know, the cost of limb isn't extraordinary? What little side hustles, what little gems can the teenagers drop over breakfast, Stephen, that you can take to an app or play around with, or you know, what's logistically really hard at home right now that you can optimize? And once you start thinking out of the box and throwing it, um, I mean, it is interesting, and it's probably a whole nother conversation too, around how the the whole tokens side of this works in terms of, you know, these subscriptions.

SPEAKER_00 13:44

It does suck you in. And so for those who um haven't played around with this too much, look tokens are basically the way that these companies work at how much it is that you're using them. So you've got a number of tokens that you can spend. It basically cuts your prompts down into you know individual tokens and sends it off and sends off the context. Everything's a token. Uh, and it's about maximizing those tokens. Um, and that's how actually they're selling it to large organizations. It's like, here's 10 million tokens, you know, for you know,$100,000 or whatever it is. And that's the way that you know the the the the Clauds, etc., are beginning to lock in these enterprises. So learning to talk in that token language um and how to use it, that's fascinating as well. There's a whole thing that's happening there.

SPEAKER_01 14:29

And that probably, weirdly enough, leads into our next thing we wanted to talk about, which was we all saw the scary headlines. The Claude's got this new model and it's too powerful, and no one can have access to it because it's gonna destroy the world. Um, and then you found that they invested, oh, what was it something like gave this to some trusted partners and we'll we'll loop back around to what it all can do and through was it X amount of millions of tokens at it going, hey So so it's actually a hundred million dollars worth of tokens that it gave to the project.

SPEAKER_00 15:01

100 million of like hey, wrote the fun park. Yeah, we're giving you that for free. So so let's so it's called Claude Mythos, eh? What a name. So it's like they're not under hypothesis.

SPEAKER_01 15:13

Stephen Fry book as well. Great Stephen Fry. Yeah, yeah, mythos. So yeah, there's there's$100 million worth of tokens because they discovered this preview, this capability. This mythos can go in and autonomously find and exploit zero-day software vulnerabilities. Massive, right? And so there was the big announcement that, hey, we've got something that is too powerful to release to the public, but we're actually going to send it out to some trusted partners, fund it, and let them get ready.

SPEAKER_00 15:43

And and and what what what that means, um in more in more layman's terms, is that this this this AI, this this LLM is so powerful, it can basically know how to hack into most computer systems around the world. That most computer systems will have these vulnerabilities. And if you know how to use this LLM and it's not that hard, it's basically just talking to it, um, it would be able to do it. And so they so they did that internally and went, oh dear, we've got something that could bring down all sorts of things, hack into all sorts of systems, you know, with not a lot of technical know-how, what are we going to do? So they they pulled the release for just now anyway.

SPEAKER_01 16:24

And it was kind of interesting too, because I know about you, but you get a bit cynical when you see these headlines. You're like, what is really going on? Like, should we why are they telling us this? Should we really know? Um, and then you realize, well, it's obviously a huge like even if it is a marketing ploy, holy proverbial, like this is in incredible functionality. And someone actually coming out and maybe linking back to the ethics, dare I say, of um anthropic, to talk about actually we're gonna try and use you know this for good. We're gonna release it uh to trusted partners to actually, you know, plug some of these holes rather than have the reverse happening where it leaks and we've got huge risks.

SPEAKER_00 17:03

This this sounds like this is going to be an ongoing issue though, because I mean these these models aren't slowing down. So is this now part of the normal? And and also, you know, this so that what they've created is this thing called Ploject Glasswing, which is that consortium of you know fifty fifty plus companies with your Apples and your Microsoft and your Googles and the rest of it, and they've brought them all in to say, hey, you need to use this tool in order to plug all the holes in all your systems before we release it. But we know that these systems aren't going to be fully patched. You know, the the number of systems you know that we still come across on a daily basis, you know, within organizations that haven't been patched for years.

SPEAKER_01 17:37

Uh all that tech debt that's hanging around.

SPEAKER_00 17:40

Yeah, right. So this is for for standard software, you know, all the all that custom-built software that's out there as well, that's going to be vulnerable by attack for this as well. So it's getting to that dangerous phase, you know, if you can use this to go and attack software that does critical things, and which is most things these days, you know, this this is powerful. This is really, really powerful technology now.

SPEAKER_01 18:03

It is. And I I think some of the examples they gave were they found, and we all of those that work in, you know, any kind of major project will be like, oh yeah, of course. We've all kind of gone into and and needed to remediate certain things that are known issues, but this is an unknown problem. They found a 27-year-old vulnerability in OpenBSD. So this was some software that has a really strong security reputation, right? And in the past, like if you put a full discovery operation team in there, it might cost you 20 grand to try and find this sort of issue. Um, but the model that found it actually cost about 50 bucks in terms of building it, right? So that's a bug that survived 27 years of human review. I mean, it's just a little bit older than me, still. Dumb it. Um, you know, it survived for 27 years and it was found autonomously by this model, right? For less than the cost of a tank of petrol. It's probably a quar probably a quarter of a tank of petrol at the moment.

SPEAKER_00 18:58

Well, that's true, yes. It's a whole other podcast, that one. And do you know who can't use this model just now or officially can't use this model?

SPEAKER_01 19:06

Go and Stephen Who.

SPEAKER_00 19:07

It's it's the US government because they've labeled anthropic to be a supply chain risk.

SPEAKER_01 19:12

So many cool moves by the US government at the moment. It's like, really? Really? I feel like this might be handy. This might have been a handy tool.

SPEAKER_00 19:20

Uh yeah, so I mean I mean, surely they are behind the scenes. You know, but officially they're not allowed to because they they can't use anthropic and claude, and they're trying to rip that out just now. How that stands. Well, we'll we'll come to that later because we did cover that a lot in the last couple of episodes. But who's on the rope at the moment, Lauren?

SPEAKER_01 19:37

Again, Stephen, with philosophically, who is it?

SPEAKER_00 19:40

Um so look, OpenAI have shut down Sora. Did you ever use Sora, the video generation thing? Did you ever go in play with it?

SPEAKER_01 19:50

No, I kind of just right from the get-go bloody hated that thing. Just as a as a muse with a lot of creative friends, like I was it kind of made me angry. I was like, oh look, guys, we'll just use all of your creative efforts. We'll, you know, every frame of these videos that was generated was basically in my opinion stallion. Um from the works of all these artists, right? So oh, so sad that we've had to shut that down. But it was actually costing a huge amount of money to run. Isn't this ironic? So sad. It was about a million dollars a day, right?

SPEAKER_00 20:21

I I know, I know. I and and it brought in revenue of 2.1 million over its whole lifespan. They tried to play that consumer market. I'm thinking what, there were the new Facebook or whatever. I don't know, right? Um, but it but it didn't play. Um one of the things that made me really hesitate and not go anywhere near it was that if you signed up to it, the mere act of you signing up meant that you in your image, it was now fair game for other people to take you and your image and use it in whatever they wanted. And I was like, really? No.

SPEAKER_01 20:55

And there's still a lot going on. You know, we could do a whole episode on this, like musicians that are you've got AI coming in and pretending to be that musician and then suing the real life musician for fraud. And oh, don't even get me started. So, not so sad, although it's interesting because you've got Claude that's come out with all this functionality, and then you've got open AI going the other way. They're scrambling to actually consolidate and come up with a better offer. So they've shut down Sora. This is six months. This is the big one, if you remember, if anyone remembers, that they signed this huge deal with Disney, right? It was like a billion dollars with Disney.

SPEAKER_00 21:28

Billion dollar licensing deal, yeah.

SPEAKER_01 21:30

And how much notice did Disney get that they were shutting it down? Less than an hour. Yes. What a Barry Crocker.

SPEAKER_00 21:40

That's just crazy. You know, was something that should be so, you know, your billion-dollar deals aren't to be sniffed at, you would have thought. Right?

SPEAKER_01 21:48

Um, but most contracts wouldn't let you do it, right?

SPEAKER_00 21:53

Yeah. Um, how they can get away with it, I I I don't know. But you know, Sora's now gone, you know, um, long live Sora. Sora. I'm sure it will live on in different guises elsewhere and might be open sourced and pushed out there, or who knows. But you know, Sora's gone, the licensing deal with Disney's gone. I think this was part of OpenAI, where basically throwing everything at the wall and just seeing what's stuck. And this one didn't stick.

SPEAKER_01 22:19

Well, it's also really interesting or possibly really boring, whichever way, which whichever floats your boat. In terms of this is how hard it is at the moment around what you need to put in these contractuals. These there's hardly any obligations around product stability or partner notification, you know, good old-fashioned escalage, those kind of things. Like we're just gonna sign a billion-dollar deal and then shut it down within an hour and then let you know. And then they've kind of also the other difficult thing they've had to face into this court is, and again, I've probably I've you experienced the ads yet? I'd be interested to know how that's coming up. So they have put in ads, right, to try and get it.

SPEAKER_00 22:57

So not worldwide. So it's only it's only certain um parts of users that that have it just now, and it also depends what tier and stuff you're on. But they're they're actively going down that path. Um because well they have to, right? Because you know, they've got such a such a huge user base there. They're burning so much money. They they did have a huge um round of raising cash very recently, but even so, they still need to bring in a massive amount of money. And you know, ads is what's done it for the last you know 20 years on on the web. You know, that's what's made, Facebook is what's made, Google, you know, uh OpenI were thinking, well, you know, that's that's gonna make us. Um maybe not though, because look, I'm the same, I'm hardly using ChatGPT at all now. I haven't left it entirely yet. I'm close. I'm close.

SPEAKER_01 23:44

Oh, it's gonna come, you know, chasing you with its little metal hand soon. And you know, I guess part of all the reason why this is all happening, right, is we've got a bunch of IPOs that are coming, right? So there's a really big all these big numbers that have been floated around, you know, that that open AI, they're targeting a trillion dollar valuation for their own IPO.

SPEAKER_00 24:06

So there's you you've got three massive ones that are coming. So you've got, you know, uh Anthropic, um, who they're talking 380, 400 billion valuation. That's a big number. But as you said, open AI, a trillion dollars. But what's potentially going to go first is SpaceX, which is expected to exceed$2 trillion. Right? We're we're into just crazy money territory.

SPEAKER_01 24:29

We think of Dr. Evil, does it make one? Oh, the world can answer for one one billion dollars. Preparation H. Oh, so mature. Sorry, Stephen, I went straight across your SpaceX with Dr. Evil. The cat. You need to bring the cats in.

SPEAKER_00 24:44

Well well, I think Doctor Evil and SpaceX and that whole analogy goes quite well, to be honest. I think it's bang on the button. Thank you. But the even if you don't care, and I know most people don't, right? But if you don't care, you've got SpaceX doing two trillion, we've got OpenAI doing a trillion, and you've got you know Anthropic doing 400 billion, right? And that's all going to happen in the next six to twelve months. What that does is it sucks out all the money for all the other organizations basically everywhere, um, that's going to these three companies instead of going elsewhere. So even if you're nothing to do with any of this, you know, what you're not seeing then is the money that's getting you know invested potentially in renewables or you know, a teddy bear company or whatever it is. Teddy bear company because I don't know why that jumped into my head, but it does right. It went from renewables to teddy bears. But right, but what's happening is that all the money is going to get sucked up this year into these, into these. And when I was reading more into this, it it sounds like because you know the institutional investors will get involved, they will by fiduciary law basically have to invest. You will see a pop in this in the in the price here. Now you may well see a huge crash after the fact, but initially, huge amount of money are going to come in here and suck out what's going to happen elsewhere. So the hype to reality, as we talked about at the beginning, you know, that is you know, that is what's happening now. This is so big, right? And it's affecting things so much worldwide now with these sorts of numbers. It's but yeah, it's absolutely transformatory, whatever way it is.

SPEAKER_01 26:28

Yeah. All these theoretical numbers around how much money it can generate, the cards are now gonna land. Um, we'll see really what it means. And you know, like you said, even if it personally doesn't affect you, we're gonna see what it does to the stock market and where this funding goes and how they actually keep evolving these products. So it's such a strange time to be living to see how this will all, you know, pan out. And some, you know, uh some of these organizations were still just figuring out where their ethics are at. Like again, you're thinking about oh, Anthropic would come out and said this. There was this huge article on Sam Altman in the New In the New Yorker, yeah. Yeah, in the New Yorker. Whoa. So and obviously we know who runs SpaceX.

SPEAKER_00 27:10

So the the the one and I know this is tiny, right? But one of the ones that came out in that New Yorker article about San Altman, because it was basically going into his character, that you basically can't trust him what he says from one moment to the next. There was a story about how Sam had said that he always wore a blue jumper all the time, or whatever colour jumper it was, and he and he said that because he he didn't have time to think about wearing a different colour jumper. And this is what he was saying to an interviewer, right? Two hours later, right? This the he sees him going into another interview and he's changed his jumper. He lied. He lied, he lied about the colour of the jumper, right?

SPEAKER_01 27:46

Well, was he stealing from Steve Jobs? I don't know. Wow. So much decision for Twig. I'm glad you figured out your jumpers. So much going on.

SPEAKER_00 27:54

But this is the guy who's good got that trillion dollars, you know, that potentially you know that that's going to be invested in his organization that he's he's driving. And as you said on the opposite, and we we dealt with Claude a lot in the last couple of episodes, but you know, a slight update in that you know, during that dispute that which is still going on with the Department of War, as we said, you know, it's now you know a supply chain risk. You know, over one million people per day were signing up for Claude. Right. Um and it became the top, the top AI app uh in over 20 countries, right? So it has you know uh and maybe you know why people were flocking to Claude is you know, if the future is AI, maybe they would prefer it not to be the Terminator.

SPEAKER_01 28:30

Maybe that's the Yeah, maybe not running the robots that are shooting at us. Yes, definitely. So it's all gonna come out to play in the next um few weeks and months, right? So, you know, are we gonna have this foundational infrastructure or AI replacing the jobs of millions of people around the world? I think look broadly sweeping in terms of you know conversations around the fireplace. I think Australia's a very different market to you know the US in terms of displacement. It's more about enabling our workforce, which is awesome, right? But let's see what happens when these IPOs go out and actually land these big claims, what it does to the stock market.

SPEAKER_00 29:09

Oh well, the thing is with the the way that the stock market is pricing this in, right? The only way to kind of justify these sorts of valuations is that you know AI not doesn't just become foundational, right? It's that it will replace the jobs and the roles. That's the only way to justify these valuations.

SPEAKER_01 29:25

Aaron Powell we unleash a bunch of talent, right? And uh enable all these new, fresh innovation and ideas.

SPEAKER_00 29:32

Well, well, we we I did send you a video uh earlier in the week, um which was which made us both laugh. And I know we shouldn't make fun of American accents but Oh, go on. I know what you want to say. Go on. So so one bank here who were um were going down that kind of efficiency pass and had let a number of people go, and there was uh an American commentator who was discussing it, and he was talking about the Ben Dingo Bank. Bendingo Bank and how how brilliant it was that uh and the Australians would call uh their bank after the name of a dog. It's Bendingo's bank.

SPEAKER_01 30:08

It was in the first minute of the video, and he had the you know, the ray bands on, he's walking down the street street calling it Ben Dingay.

SPEAKER_00 30:16

Yeah. Um context is everything. It it it's bendigo for anyone who isn't from Australia, it's bendigo bank. Go back. Um anyway, I thought that was a cracker. Okay. So look, there's so much extra you know, we could we can do we've we've touched the surface on you know a lot. Hopefully, this has you know sparked some interest for people out there. Um is this rope a dope? You know, is you know, has has Claude been taking all these, you know, Anthropic taking all these punches on the rope? Um you know you know Mohammed Ali didn't win in Kinshasa by being passive, he he won because he had a plan and and Foreman didn't. You know, has has OpenAI just been, as I've said, you know, throwing stuff at the wall and seeing what stuck, whereas actually Anthropic did have a plan? So you know, maybe that whole constitutional challenge and you know, and restricting model releases and the enterprise focus, maybe that's the strategy that's going to work its way. Or, you know, um you know, and opening eyes and you've cutting certain things down. What do you think, Lauren, overall?

SPEAKER_01 31:14

Look, overall, I think it's important if we're reflecting on Ropadope, Stephen. Um, did you hear that you know, during some of those fights that Muhammad Ali burnt his hands a few times?

SPEAKER_00 31:24

Did he? No.

SPEAKER_01 31:25

Yeah, that's what happens when you punch George Foreman in the grill.

SPEAKER_00 31:29

Oh dear. Ah, see what I did.

SPEAKER_01 31:33

I loved it. Terrible.

SPEAKER_00 31:37

You didn't get that from ChatGPT, did you?

SPEAKER_01 31:39

No, that wow. No. No, I must I must confess he used an old friend, good old Google, and then riffed on it. I love a pun. I love a pun. Excellent. Um, let's see how this all pans out in the next few weeks, how the IPOs go. Go and check out that Sam Altman article. But um, yeah.

SPEAKER_00 31:56

Amazing. Thank you again, Lauren. We shall be back soon. Um, and look, thank you everyone. If you you know, if this has got you thinking, please share it with someone wrestling with these questions. You know, subscribe and comment, you know, wherever you're listening, it really helps. And we shall see you next time. And remember, you still matter, at least for just now. At least for just now. Thank you, Lauren.

unknown 32:18

Bye.