E533 | This Week in European Tech with Dan, Mads, Lomax & Andrew Scott

28:19Well, look, I mean, just not to do the UK down or the EU down entirely. I mean, they have come up with plans in relation to this, right? And also, to be fair, just to interrupt you very rudely, this is whatever. Last year, this can change. We can have new plans. Things can move. I mean, things can evolve. Maybe it's not fair of me to compare like for like. Maybe we're not comparing apples for apples and we'll never be the crazy sales go rah rah Americans. We're not. We're Brits. So we're Europeans. So maybe we're not comparing apples and apples. And these things can absolutely change over time.

28:58But it is I mean, it is important to remember that this this plan is is the product that was triggered by an executive order that Donald Trump signed when he first came into office. Basically, that said that, you know, make the U.S. the dominant dominant power in A.I., right? So this is the sort of immediate follow on from that. And this does read very much like a manifesto. And probably, you know, everyone's talking about investing in infrastructure and sovereignty, et cetera, et cetera. I think, yeah, as Andrew alluded to, the biggest thing in the US report is just how stark the deregulatory drive is.

29:33You know, it does read like straight out of Silicon Valley. It's hand in glove with the private sector. you can tell David Sachs, well, he's a co-author of the report, you compare that to the UK, which is much more considered. The US will actually penalise states that get in the way. And we talked here about it before, but there are data centre projects in the UK that are held up because they can't get power until 2035. So there is a stark difference. What could we do more? I mean, look, the UK has the plan. The UK is investing. Last year, the UK cancelled the Edinburgh supercomputer. This year, they finally saw sense and are, you know, approving it.

30:16And that's going ahead. So, you know, this sort of to the extent that the politicians here have the capability and have the money, they are trying to do that. I think we should mandate. But they can't evolve more money. They can't evolve more money, Andrew, because that's just too obvious. They can't involve money. You have to suck up the pain and take your money from somewhere else. I think we should mandate a national AI compute and data backbone. So that sort of means like a sort of AI commons where there is funding to include national GPU cluster, open training of data sets, which are accessible to public institutions and startups.

30:54you know make sure there is guaranteed public investment for that over the next five ten years so that the next government can't reverse it you know Europe's been doing the same with this they've been sort of tiptoeing around the sort of by European by UK I think we just should make it a legal requirement that any public sector has to buy UK first if capabilities are equivalent thirdly people always talk about talent but we can be tip down tiptoe around the edges so you could have a sort of AI talent sovereignty act where, I don't know, you fund, you fully fund 5 ,000 PhDs in AI and you have a special AI tech visa, which is, you know, which is authorized in 24 hours.

31:39And you set tangible goals and you throw money at the problem so that we stop bleeding talent externally and we start pulling talent internally. And Europe could do something similar. I mean, there are three things just off the top of my head which would have a tangible move the needle. Might not be the best things, but we need to start making decisions. Margaret Thatcher said, the most important thing is any decision is better than no decision. And we just don't, we keep not making decisions. You know, you mentioned the computer, the supercomputer thing, it was canceled and then regurgitated almost exactly the same 18 months later.

32:11Yeah, no, you're totally right. I think to be fair, on the infrastructure, they get it and they're doing it. So they're pushing for a 20x expansion of the AI research resource. Right. And that actually is happening. The money's going in. So on the first point that's happening, the second one is very interesting. You know, some people will, will jump up and down about protectionism, but I think we just, you know, squash them. And the third one on talent is absolutely, absolutely needed. By the way, let's not forget our, our European friends. They have this, they have this kind of push the Euro stack, they call it to basically drive European tech sovereignty.

32:45And they have 300 billion AMARP to implement this, yada, yada, yada. But, you know, there are these initiatives going on. But, you know, I think maybe consensus here, they might be too little too late. You need like an AI Marshall Plan. You need something a bit more dramatic. Because otherwise what's going to happen is I think for a lot of countries. Steps ironically not written by an American. Not written by an American. Call it a Montgomery Plan. There you go. Call it a Montgomery Plan. The Churchill Plan. Yeah. The Churchill Plan. Even better. We can drag him back up. I think one of the challenges will be.

33:18Not the Keir Starmer Plan. No, not the Keir Starmer plan. Very sensible, very calm and peaceful and sensible. Take some time and consider. A lot of the US providers are getting wise to this alarm in Europe. So they're starting to offer sort of, what do they call it, sovereignty washing. So they're sort of saying, hey, it's fine. You can still use Microsoft Cloud, but we'll make sure that it's, you know, on premise, domiciled, whatever. It's ring fenced. but actually that doesn't solve the problem because it's still all controlled by the u.s you know ultimately it's u.s technology so it doesn't really solve the problem but i suspect i suspect that's what a lot of people are going to do i saw microsoft amazon i know datadog snowflake and others are all building data centers or running local environments across europe and investing billions and billions so there is some upside to that uh ring that regulatory ring fence that the EU creates.

34:13How much that's going to move the needle? I don't know. But there's one other point that I wanted to make before we move on, which is around the... I was reading this week about Norway's Sovereign Wealth Fund and how they applied CLAWD about 12 months ago. So the Sovereign Wealth Fund is run by a guy called Nikolai Tangent, I think. I don't know how you pronounce it, but Nikolai Tangent. And they managed$2 trillion. $2 trillion is quite hard to get your head around how much money that is. It's so many zeros. I just don't think I think it becomes quite nonsensical in the human brain. But that is a vast sovereign wealth fund.

34:47And they've seen a 20 % efficiency gain by applying Claude to manage their back office operations, solve bias, highlight investment strategies and more. And they only employ 670 people. What he says to his staff is if you don't apply AI, you will not be promoted. So that's how much they he's a proper, proper fan. And they've seen some massive, massive benefits to applying AI early. I mean, they were doing this a year plus ago. So I think that's a nice, interesting kind of lens to look at how large orgs or quasi-government, these big kind of money organizations can be very smart with AI in large operations.

35:27I think it's interesting. And I think that often you see tech CEOs who are mandating this saying, if you don't use AI, you're out the door. It's actually interesting to see a government organization or Southern Worth Fund do that. By the way, the final reflection on this U.S. plan, we've really come 180, right? Do you remember when, remember the, you know, we know about, you know, move fast and break things, right? This sort of U.S. ethos. You know, when we had under recent governments has been lambasted and we've obviously seen big tech CEOs dragged through the Senate on congressional hearings, et cetera.

35:58like this document that they just put out the one with the big eagle on it that is literally like move fast and break things mandated by the government as mad says forget copyright forget like water regulations forget everything just build baby build as they bloody said they've even got marco rubio saying build baby build now i mean that guy is now he's on he's uh he's i'm not convinced i care what he says i think i think he can say whatever the hell he likes but We can't be the Apple of big tech when it comes to AI. I think we are all in complete agreement with. This is the one technology, because what Apple would do is they would wait for Samsung and or Google, whatever, to build.

36:45And they would apply their version in two years. And obviously, they've been caught with their pants down this time. And the UK government and the UK cannot be Apple in this big tech sandwich because we know how badly that ends. So let's move on from AI. Let's talk about stablecoins. The UK is just now starting to think about stablecoins. And so what you might say, what is a stablecoin? What do they do? How will they work? How will they affect in the financial system? Will they create opportunities? So for those that don't know, a stablecoin is a cryptocurrency pegged to fiat. So it's pegged, i.e.

37:17to the dollar, like Tether or Circle or TrueUSD to reduce volatility, enable smarter payment rails for e-com, for cross-border remittance, micropayments, DeFi applications, asset tokenization, and even treasury potentially. So this is absolutely unlocking the power of blockchain, much further beyond the Ponzi crypto nonsense that we've seen not long of all, though that's actually still going. The EU has the MICA, which is the Markets in Crypto Assets Regulation Act. The US has passed legislation recently. They passed the Genius Act. Now, this looks at stable coins and make sure that the rules work around how stable coins work.

37:58And there's obviously nothing wrong with thinking about these things and smart legislation and creating guardrails, regardless of whether you do anything with them as a country or anywhere in your financial system. Paul Marshall, I noted this week in the FT, noted down that Japan introduced a stable coin law in 22, Singapore in 23, Abu Dhabi in 24, and the US just recently with their act, actually this month. And why are we just starting a consultation process now in the UK? So stablecoins, the two dig in, low max, useful, meaningful, kick off. Well, I think maybe I'll take this in the direction of what I think this is a big miss for the UK.

38:39The financial services sector in the UK is actually its crown jewel. it employs 2 million people delivers 100 billion of tax to the to the revenue in the uk which is roughly 10 of the tax take you know historically the footsie 100 was one of the the best biggest most prestigious stock markets in the world so it is quite frankly shocking that the uk has actually been i would say out to lunch on this completely and has pretty much shut its shut its ears and eyes to this and has allowed other countries to move forward and start to establish the regulatory regimes which provide the certainty and the seeds for you know this innovation to take root in those economies so i think the uk has been caught flat-footed shockingly a week ago Rachel Reeves was in Leeds giving a speech which have now been called the Leeds reforms which were all around making deregulating and making the financial services sector faster nimbler bigger less encumbered by red tape and they've completely missed what is potentially a critical innovation within financial services is there some kind of risk that they are concerned about and therefore putting their fingers in there is why yes yes yes their risk about financial stability so Andrew Bailey actually who's the governor of the Bank of England who actually is interestingly enough has just been appointed to a global very important oversight board called the Financial Stability Board and which which may be what he's maybe what he's obsessed with stability right he's now about to become the the global policeman of financial stability but um he is obsessed around risk so yes and and he's worried about that and i don't i have no doubt there will be risks around that but i think you've got to have a horse in the race i think you gotta have skin in the game and i think you've got to be involved and not shut your eyes and ears to this kind of thing particularly post-brexit where the uk does have some flexibility here you know people had talked about singapore on thames and a bit more freedom to operate etc etc It seems in this case, like a few other cases of late, this is a mixed opportunity to start to take part.

41:07It's taking part in the conversation, I think. Still, this is early days. And I think just ignoring it is a risk that the UK gets left behind. Mads, one of the thought that I had was this is just a way of the US extending the life of the dollar. Is that too cynical or is there some merit in there? I'd say absolutely. It's about dollar life extension. But I think it's more sophisticated than simple currency dominance. But I think you talked about some of the numbers. You've got more than$250 billion of stable coins globally. 99 % are over the USD, US dollar denominated. They're offshore today. And the Genius Act that passed now will force all of this onshore within the next three years.

41:50And that means you have basically a new system of dollar dominance that is then solidified in the stablecoins. And the more transactions flow through these stablecoins, the more the U.S. will gain, should we say, the lock on the next generation of the global financial system. So I think it's very smart. I think it's the right thing for them and their economy. And I think it will bypass SWIFT, but it will strengthen dollars. Now, China, they clearly see this as a risk from a sovereignty perspective, and which is why they banned crypto, but they've been pushing their own digital yuan, their own digital currency.

42:28So I think the US Genius Act, which kind of mandates, I think, Lovac, you touched briefly on this, it mandates a one-to-one dollar reserve. So that means that for every time you create one of these stable coins, you have to hold the same number of dollars on the balance sheet, and that basically forces those holders of the dollars to purchase treasuries and that's where the 100 billion or more comes from. So it becomes a massive source of cash for the US treasury and that will give network effects, they'll compound over time. I think we have been out to lunch, we've been sleeping while this happened.

43:07There's sort of a bit of British institutional paralysis there and it's a shame. I mean, there would have been no reason why London could not be faster here. A historical parallel is perhaps the adoption of electronic trading back in the 80s. At the time, London nearly lost market leadership because we were slow to react. The challenge we have this time is we don't have a leadership to lose. We are already behind. So we just can't afford to be asleep when you're from behind and you're trying to accelerate. So still lots to work out. But I think that the Americans have made a smart play here. Yeah, no, it feels smart, especially as we talked recently about how much is going into gold and obviously into standard crypto.

43:49So it feels and BRICS is coming online and there's less dollar interaction and with the power of tariffs. So it feels like this is a nice play for the Americans to try and get more dollar throughput through this stable coin mechanism. Andrew, would you add anything or how do you see this world? I just think it's such a shame because the UK stole a march with the rules around banking and the FCA really embraced digital banking. And that's why ultimately we have a couple of very large, successful fintech winners. We're at this crossroads where I don't think it's quite too late, but Mads is absolutely right.

44:31It very nearly is where we can choose a sort of state led top down, strong central regulation, limited private issuance, you know, back to cautious innovation of Europe and China. or we can take a sort of Singapore slash UAE approach, which is pragmatic regulation, fast, simple frameworks that kind of balance innovation with control. And the US is tipped to being probably more ahead than that. It's kind of market-led and private stable coins and innovation first, much like the AI. So there's a real cultural theme here. And it's just a great shame that, as Lomax says, the finance sector is something where we have such a long history and tradition of market leadership and we're stumbling on it but it's it's a ideological issue and i don't expect labor to want to give up that central control and take a radical what they see as a radical step and kick the fca up the bum who are naturally extremely risk averse you know and rules first So heavy on consultations, lack of action again.

45:41If we wanted to take a, should we say, more ambitious approach, something I think could be done. I mean, one of the concessions that was made in the US legislation was around interest and the ability to pay interest on these stablecoins. And the legislation basically bans the payment of interest on stablecoins because there was a fierce lobbying from regional banks saying, well, if you get interest on stablecoins, people will stop putting money into our small regional banks. That's competition. That's what competition is. Well, banking, as you know, have had a chokehold on provision of credits for centuries, and they do not want to give it up.

46:22Well, this is an area where we think we could steal a march on the Americans, because if the UK could be the place where you have not just stablecoins, but stable coins that are interest-bearing, it would change the game overnight. And it's that kind of aggressive innovation, I think, that is required now to try and come from behind and see if we can get back on the game. This is possibly an unanswerable question, but I'm intrigued how in the real world stable coins are going to be used, what applications we're going to see that are going to be meaningful, how investors and or startups will capitalize, let's say the UK gets its act together and we understand, we regulate, we correct the guardrails around stablecoins.

47:09How does this layer down into our worlds, I guess, is my slightly mealy-mouthed question. No, I think it enables on-chain banking, the ability to use it in decentralized finance. It often avoids currency conversions. It releases control from the banks, your money going through a central source and this trusted third party provider. So none of them like it because it ultimately ends in a challenge to the traditional structure of banking, which is a trusted third party controls the money and takes a handsome reward for doing so. So that's why they're all against it. They will argue the point of regional interest and regional affecting these different things, etc., etc.

47:57But ultimately, it's an existential risk to the existing financial system. I get that it's Rails. I get that it's DeFi. I get that it's cross-border micropayments. I'm just wondering how much that is worth. I would say stable coins are a bit like digital money that moves more like email. Who are going to be the winners and the losers there, I guess, was my unpack. But maybe this is a conversation for another time. I was trying to work out what the real impact of the UK getting its stable coins regulation together. What's happened in Singapore since they closed their regulation? Or in Abu Dhabi?

48:33Or what will happen in the States? So maybe it's no bad thing to see what the financial outcomes will be and then follow the best of breed as to whoever did it best in the world. Maybe that's not a bad thing, just to be upside for a second. You know, as you say, and as we said, the U.S. is already, you know, over 99 % of all stable coins are over the U.S. dollar denominated. And we saw the Circle IPO last month perform phenomenally well. And what's happened now is that the infrastructure companies of tomorrow, right, last generation was Visa and MasterCard and it was Swift and it was all those legacy players.

49:14And the parallels of that for tomorrow will be Circle. That's why the Circle IPO has performed so well, because that will become a fundamental pillar of the global financial system of tomorrow. And wouldn't it be great if one of those pillars was headquartered in London instead of the United States? And we have incredible founders in this space. If you look at fintech, you look at Klana, you look at Revolut, amazing fintech entrepreneurs. Norse. No reason that Europe could not build one of these next-gen players. Okay. Well, let's see. I'm super intrigued as to how this is going to unfold. It's also, Dan, the whole industry that thrives off it, right?

49:54It's the lawyers, it's the bankers, it's the consultants as well. Those are big drivers of the economy. Banks, credit banks, credit card networks, the regulators lose out, any sort of high-fee remittance providers like Western Union, you know, a whole layer of control which causes friction for what is ultimately just my ability to barter sheep, right, gets removed and accelerates the ability for people to trade, build their businesses and make money. And if you believe that fundamentally a job of the government, as well as keeping everyone safe, should be making it frictionless to trade and build and grow and therefore grow the economy and therefore make a better society, then you want to remove friction.

50:42And this again is, you know, it's the tail wagging the dog. We've got all the people who create the friction, arguing that we should create more friction, rather than removing friction from the real economy. And the real economy is the people that make things and deliver services. All the people that look after your money and take a cut from you transferring money, they're all bottom feeders, right? They're not making anything. All they're doing is just allowing you to do what you should be able to do really easily. I totally get it. I totally agree. I think that it will be interesting to see what monetary value this creates.

51:17But like I said at the top of the piece, it kind of doesn't matter because we should have this agreed anyway, regardless of all of these potential outcomes. We should just get our act together to work out how the city is going to deal with it. So I think it feels like hopefully this will give them the kick and these things will happen, but we'll see. Changing tack, I want to talk about how Google and OpenAI are good at maths. And there is a reason for this. So Google won gold at this maths competition. It's the IMO, an OpenAI who weren't actually officially in the competition claimed to win two.

51:53They won gold medal scores at this IMO, which is the International Mathematical Olympiad. Normally, this is humans. There are 630 odd students that go into this contest every year. This was the 66th down in Australia, where around 67, but 11 % of these students win gold medals. And it's a hyper, super, super maths elite competition. Last year, Google's DeepMind only won a silver. This year, they won gold. So what, kind of you could say, you know, every year or every week we hear a new AI winning, AI doing story, but we are potentially on the path for AI to solve extremely difficult frontier maths problems.

52:37Stuff that we've never, ever been able to solve. Even these 11 % of these hyper brains haven't been able to solve. What does this mean beyond the maths competitions? What's the real story behind the story? Well, it means that AI is crossing the critical reasoning threshold. And it's not just computation. It's about creative problem solving. It's about proof construction. It's about mathematical intuition. Now, last year, DeepMind was at the silver level. And this year, both Google and with the DeepMind unit and OpenAI performed to gold level standards. OpenAI didn't enter the competition formally, but showed that they could deliver to the same output.

53:22If you look at the competition itself, so there are 630 humans that compete and about 11 % won gold. And these are the best and the brightest in the world. So that means that AI systems are matching the top 11 % of the world's best young mathematicians in performance. It's an absolutely key milestone. And it represents a leap from just following patterns to actually inventing solutions. Now, these models were not specifically trained for the Math Olympiad. This is emergent reasoning, which is why it's so exciting. And if you think about where math applies, well, it's the foundation for scientific discovery, for physics simulation, for drug discovery, molecular optimization, engineering.

54:09so many areas where this could be used. Now, some of the historical parallels we're looking at when computers beat humans at chess in 97, the skeptics, they said, yes, sure, but it can't be done in Go. And then when AI beat the Go champion in 2016, people said, well, sure, but it can't really perform that well at creative tasks. And now AI is really crossing into the pure reasoning domain, which is sort of the last human fortress. Now, if we look at over the last few years, the evolution of this, 2019, GPT-2 struggled with basic arithmetic. By 23, GPT-4 could handle algebra. And now this year, two years later, there's a gold medal, which suggests that by 2027, you're going to have advanced reasoning that can surpass mathematical frontiers and go places where humans have not been able to go yet.

55:10So this is very, very exciting. It's got implications for corporate R &D, for material science, for financial modeling, for logistics. So many opportunities here. We often say, what's the link back to Europe? I mean, Europe's got some incredible traditions in maths, whether that's out of the French mathematical schools, whether that's Germans with their Max Planck Institute or Cambridge and Oxford in the UK with some very, very strong mathematical heritage. But obviously, academic excellence is not the same as having strong commercial application. And so the question I'd ask back to you in the room is, what can we do here?

55:54Should we have a sovereign response out of Europe? Should we create a CERN for AI? where we pull the best and the brightest to develop this next level of mathematical reasoning, then how can we take advantage of some of these advances to benefit our economies? I'm more worried about the end of humans. But Andrew, do you have any thoughts on what we should do in Europe and the UK? Yeah, I mean, I think it goes back to the previous conversation about AI in general. I mean, these models are super impressive, but they're still sort of statistical juggernauts, right? They're not really generating the solutions in that LLMs sort of don't know what they know, right?

56:37They don't understand concepts in the way humans do. And I'm still of the more skeptical end. I think they'll continue to impress us at great speed over the coming months and couple of years. But unless I've got a fundamental misunderstanding, they're still optimizing probabilities of tokens, right? the next token it's it's not it misses the human self-reflection the curiosity the the value alignment that i think creates great science and so that that for me it is for me more of a chess you know great you know what was it called ibm thingy blue that did the blue is more deep blue thank you thank you see you see the llm wouldn't have forgotten that name would it it would have known whereas my mark one the human brain forgot it is a bit more like a you know a mass version deep blue for me i don't think it demonstrates that however clever it is that somehow we're seeing a shift to more human levels of thinking mads anything that you close off on here before we move on to deal of the week anything else you close off i sort of come a little bit back to you know is this just a clever parrot that's parodying stuff it's seen before and it seems to me that every time we hook who the models we do so at our peril i go back to the chest to go go to creative.

57:56I go back to, you know, we've talked about strawberry. There's a fundamental reason why LLMs that operate at the scale of tokens are starting to see what's inside each word. But obviously that's a problem that's been solved as well. Now, and every time we say this can't possibly be done, well, six months later, somebody's done it. And I think we are on the verge of some really, really interesting breakthroughs here, also scientifically, and it will transform the way the world works. I mean, in many ways it already has, but the first time we see a scientific breakthrough that comes out of research team that leverages llm to create this and am wants to create this you know i think it's going to blow our heads off and i don't think we're far away i'm with you andrew i'm not with you on this one i think there is there is i keep being absolutely blown sideways by by what i've always perceived to be the greatest auto complete in the world um i i am i am constantly surprised but let's move on to deal of the week It's amazing.

58:54It's amazing. I'm not saying I won't be impressed. I continue to be impressed, but I'm saying it's just it's not thinking. It just it looks it's very clever at looking like it's thinking. OK, yeah, I don't think I can win this one. And it's not done that. It's not done that to the level of the top 11 percent of the best and brightest young mathematicians in the world. That's how good it is. It's looking like it's thinking. I mean, it's it's a staggering, staggering achievement. It almost doesn't matter that it might be synthetic looking like. In my head, it's like, if you're going to solve this molecule for cancer, I don't care if you're a synthetic.

59:31I don't care how you work. It's that you've created this new way of solving or doing or building or working. It's almost like a red pill versus blue pill moment. I mean, do we really truly exist or are we just living in a matrix? Do humans really truly have independent thought We all just big probabilistic machines responding to the paddles and stimuli of our nervous system. I want to talk deal of the week. Lomax, this is one of yours. This is Eurasio, if I've said that correctly. Over to you, Squire. Well, it's not a deal of the week. It's a fundraising of the week, especially in the context last week of us talking about the European VC funding commitments being at an all-time low, but certainly the lowest it's been in the last 10 years.

1:00:1610 decades. In a decade, yeah. In the last decade. So it's interesting to see this is a Parisian fund, based fund, Eurozero, who have closed a 650 million fund to support AI First European scale-ups. Good for them. I think it's interesting because also it's not only a big fund, it's also scale-up capital and not just seed capital, which, you know, to be honest, there's, you know, we're kind of all right for pre-seed and seed capital in the big picture. We are very short on scale-up capital. So that's great to see. Matt, do you want to talk a little bit about the USEU trade deal? No, we're just talking about deals.

1:00:54And Trump's erratic approach got a lot of flack. And I think in many ways, rightly so. I think it is interesting to see that he's managed to get a deal in place with Japan. I think that was probably a hard-fought deal. And I think the rumored deal with the European Union is obviously the prize for him. It's always been the big buck bear for Trump. And so if he can get away with a 15 % one-sided tariff plus quotas on European imports, actually, I think he would have done quite well. I think that's more than he could have hoped for setting out, or at least he would probably have been hoping for the world.

1:01:29But that seems like a very generous concession from the European side. And so, yeah, I'd say if the US can pull that off, that would be my deal of the week. That's where Japan ended, wasn't it? Didn't they end on 15 % tariffs generally? he has wanted more from the eu for sure and i'm not saying that the deal would be good for europe i'd say it would be good deal for the us okay okay i think japan was 15 by the way dan yeah it was i'm sure i didn't do a deal of the week this week but i'll tell you something funny about ai if you like i've i've been i've been looking for weird and wonderful uses of ai and one of the funniest ones i found was a startup called ground not a girlfriend using computer vision to prevent raccoons breaking into houses, which I thought was a niche use of AI.

1:02:15It was great. There's a German city that's overrun with raccoons, American raccoons, so they're spreading across Europe. So this could be a billion-dollar startup opportunity. Billion-dollar raccoons. There you go. I did see, I've been reading a lot, and maybe it's because my brain is now wired to look for it about people having relationships with LLMs and best friends with LLMs, and that was so that's been attracting my eyeballs this week but kind of scares me but gentlemen all the loves thank you so much you have been listening to Upside and we'll catch you next week

1:02:52Tear down this wall It's more than just an alliance This is a union of values Let's start acting