Is AI 'Safety' Talk Really a Regulatory Moat Against Cheap Chinese Models?

Transcript

If all of a sudden we put all of the agents thinking into like how can we get you into a tiny cube that's where it gets problematic, right? Because now the agents are in a tiny cube and the cubes can go anywhere, etc., etc. So the the issue that we have is like we have spent 10,000 years really slowly fumbling and bumbling our way towards glory and now we've invented a thing that might spend like 10 years and be way smarter than us. Hey everyone, I'm Kar and welcome to Unusing Money because what happens onchain never stays on chain. Before we begin, here is a word from the sponsors that make the show possible.

>> This episode is brought to you by 1 in Aqua, the shared liquidity layer from 1 in. Back multiple liquidity positions with one wallet balance and keep your tokens in your wallet until a swap fills. See how it works at 1in.com/aqua. Hey everyone, I'm here with my co-host Taylor Mahan, security expert. And we have a very special guest. We've been trying to get this guy on the show for like 10 weeks and he keeps uh keeps ghosting us. So, I'm f I'm glad we finally got you, Tommy. Uh founding partner of Deli Ventures. Uh very good to see you. We haven't we haven't caught up in in person in a while either, so uh sounds like you're you're a bit busy at the moment.

>> Yeah. Yeah. New new baby. Uh Kane, great to see you. And Taylor, thank you as well for all that you do, for the security in the space. I follow you pretty closely on EX. So, thanks for everything you're doing. >> Wow. >> No, thanks for joining us today. And uh yeah, tell us, how's life with a new baby? >> It's uh it's it's amazing. I'd love to sleep more. I'd love a couple uninterrupted hours. >> Don't worry. I've got a 10-year-old and she still doesn't sleep. So, uh you got you got you got a decade ahead of of >> Jesus. Um, >> sometimes I feel like we could just pick them up when they start pouring memories, you know? Is that too dystopian or is that too >> they still don't sleep then?

Unfortunately, that's the problem is uh is you think like, oh, like two, three, five years. No, no. >> My 8-year-old, she she wakes me up like but then she's like, wait, actually, that's a terrible idea. Don't wake mom up. And then she just steals my phone and watches YouTube slow. And so at like 6:00 a.m. I'm like, "What is happening? Why >> Why are you still awake? >> Why am I having these weird >> Why'd you ruin my algo?" >> Yeah. Yeah, that too. >> Yeah. >> Yeah.

>> All right, let's let's get into it. Um, clarity clarity is dead. Uh, we didn't even get to 50 votes. I saw a lot of people uh in in different channels on the timeline like nashing teeth being like we spent I don't know like people were like we spent a hundred million dollars on this [ __ ] and they didn't even get to 50 votes. Um, so yeah. Uh, I mean it's it's funny, right? Because I think like this is s such a cryptoesque thing, right? Like three days ago pretty much everyone would have said to you there's no chance that this thing it's dead, like it's over, whatever. Like

there was a few people that I knew that like were a little like more connected that seemed to like be kind of hinting like ah there's a chance or whatever and but most people were like ah no there's no chance it's dead right and then we had this like >> whiff of hope opium that just got like injected in the room and everyone's like ah it might be fine and no it wasn't fine. So yeah what's what's your take? Uh what's your take on this Tommy? >> Uh I I mean I think it's insane. I mean, we're wheeling in like 80-year-old senators to vote on something they don't understand with absolutely no context of how this affects the future. So, I think it's just ridiculously dumb that we trust these people to determine our future. Uh, so I think we should vote

out whoever voted against it. I mean, that's just like step one. Uh, step two, I mean, it looks like based on the SEC and the CFTC comments that they're about to unleash as much uh enablement as they possibly can. I don't know what that means or or what they do. >> Um not a lawyer, but it seems like they're just going to go nuts making sure that crypto doesn't go away. >> Well, you know, it's it's interesting, right? Because like you have the the things that we don't like, right? Um like the SEC when they're making bad rules and now we're like, well, at least the SEC is allowed to make any arbitrary nonsense that they want.

>> So, we're kind of saved by that. And it's like, man, like be careful what you wish for. like I'll I'm here for I'm here for the SEC making crazy pro- crypto rules. We we need to uh balance the ledger a little bit clearly. Um but like that's not the most sustainable strategy for us realistically. Um you know relying on on uh the SEC rule makers to to come in and and you know do stuff. I mean it is it's like it's harder I guess uh for for you know another regime to come in and like you know completely tear that down but like this regime came in and like stopped the SEC completely

from doing anything let alone making rules right so you know again I I just I do worry that like we had a chance there was a chance to actually get some laws written and we've somehow squandered that chance. I don't know. Yeah. I don't know if you guys have takes on like what what do we do wrong here? Like what what's the >> let's Monday morning quarterback this thing. Like what what should what should we have? >> I mean I don't know how this world works at all. I feel like it just lost momentum though, right? Cuz like I feel like a few months back or like when genius was happening, right?

>> It was like we there was like real momentum there and like progress and like conversations constantly. There was back and forth. that there were people um ne like negotiating is the right word but like there were iterations happening and there was discussions there was debate like there was you know there was things that were happening here I feel like I don't know I kind of forgot about clarity when it came back like and I don't I I feel like that whenever that happens it's you're never in a good position. I've also like in the last few days seen a lot more conversation on like what was included in the bill, what was ripped out of the bill, what was uh edited I guess a lot and

you know I guess you I apparently you have to make concessions when you're negotiating. However, I think some of the things that happened were a bit odd. like some of the the the the some of the things that were edited are just like full full court deleted like were not I don't know weren't the best. Let's >> The harder part though is it seems like there was like major ethics debates around provisions and giving stuff up and it like not a lawyer. I didn't even read the bill. Who does? But it seems like there was like a massive concession at one point and then again this Sunday and it still didn't pass. Which makes me feel like there's got to be some bad

faith elements to this. >> The ethics parts of it were I think the biggest was why it sort of got blocked early on. Why it sort of like started to falter early on and then and then people were like, "Okay, well, okay, we can make some concessions there or whatever. But I'm not actually sure if the concessions that were made were actually actually addressed the ethics concerns. And I think that's like the biggest I think honestly that's in my opinion not being a very political person, not not being a side chooser at least like I think that's the >> honestly I think Trump and Melania and the shitcoins and world liberty I think

all of that >> amazing. Yeah. I think that's what did it in because like it just leaves a bad taste in your mouth. And again like I >> do do you guys remember when the the Trump coin launched? Like I remember sitting on my couch. It was like late at night and I'm like this has to be fake and it was during like the crypto summit in DC. So like everyone in crypto didn't even >> was there. I know. >> No, I remember I was texting >> at the time I was still working for MetaMask and I was texting our policy guy. I was like, "Bill, what the hell is going on?" He's like, "This is wild."

That was his response. And I was like, "Oh my god, no." He did not say, "This is fake." He said, "This is wild." And I was like, "Oh, we're done. We're done for, >> dude. We're freaking done for." Like >> it's it's so it's it's so crazy, right? Like so many crazy things have happened in the last two years that you would have said like impossible. Like 0% chance, you know? It's it's a less than tail risk, right? But like things that you would have said would have been the most bullish thing possible has turned bearish on us, right? Like it's like this will never happen. If it did, it would be the most bullish thing ever. It then happens and it's the most bearish thing ever. And and >> yeah,

>> because I mean Kate, if you were >> like Kate, if you were if you were starting like a new project in DeFi, like would you rather do it now or like five years ago? Like what would you know what I It's confusing cuz like that was cyber buck era. >> But >> dude, okay, I'll tell you I'll tell you one thing. I'll tell you one thing. I would rather do it now because there is like 100% less likelihood that a [ __ ] crackhead lunatic is going to come and kick my door down and shoot me in the head. Right. I'll still take that. Like I'm talking about I'm talking about getting >> I thought you were talking about like Hunter Biden.

It's kind of what I thought. >> Yeah. I don't know. Is he allowed to leave the country? I don't know if he's allowed to leave the country. >> So, I'm probably okay in Australia on that one. But no, like genuinely genuinely, right? Like like it's pretty clear. It's pretty clear at the moment in the current regime. And this is where like you know tying yourself to regimes is like not the brightest idea. That said, if the current regime that you are in going back, you know, three, four years ago, is trying to kill you, then pretty much any regime is better than that, right?

So, like you probably should pick a side. And so, we did. And like the current regime is not trying to kill us, but like also doing a lot of dumb things to us, right? So, you know, um it's it's a mixed bag, but it's better than having your door [ __ ] kicked down by a bunch of goons and having your bl your brains blown out, right? So, so um I'll take it. Like, I would far I would far rather do that. Now, the flip side is like what's the back stop to that, right? like you know so so um okay you uh you were you know in an environment where um it was you far less competitive um you know there there was like uh less

certainty around like would it even work or whatever. So like the the you know um it's a higher variance outcome if you pull it off five years ago right than it is today. Like today, you turn up and you're competing against like listed entities on the NASDAQ that are trying to do the same thing as you. Like it's a much more competitive environment for sure, which is good. That's good for crypto. We want good strong competition to deliver the best products, right? Um it was not that competitive, you know, like in in 2018, 2019, 2020 even, right? Um so I think there is a trade-off there for sure. Uh but but you know knowing that you're not going to you know get sued to oblivion uh at least in the short term

for the next couple of years um I think is one of the reasons why we have better competition you know like there's better better people have come into the space and gone ah I can do this thing right especially especially try so that's I think that's the trade-off space in my mind >> yeah I do think we have a lot more clarity today in general >> versus no pun intended >> no pun intended But especially you go back like three years, I think it was just >> it's not only that you didn't have clarity. You actually did have a certain amount of clarity, but that clarity was like deep uncertainty.

>> Yeah. >> They might like you, it was very clear that they might come after you for arbitrary reasons if you were, you know, anywhere in the US or anywhere within reach. Um, and I think that was super harmful. I mean >> the crazy part is like we would hop on calls with like brilliant founders like early days of the venture fund and we would talk to them and these would be like young hungry guys and girls wanted to change the world and it was so interesting and like so exciting and we couldn't wait to back them and then on the other hand it's like but they might go to prison like it it just so weird at that time >> Gensler so weird it was so weird um yeah it was not the right way to do it and

and you know And that's that's why I say like right like you know uh kind of like rule making bodies right like setting setting the agenda for an entire industry is a very fragile equilibrium right like it was fragile when they were trying to kill us now they're not trying to that's also fragile so you know I I do think this is like a huge missed opportunity um and you know I know a lot of people worked really hard to try and get so the line um I think you know Tommy to your Right? Like the only, you know, the thing that I think gave us some hope of passing clarity realistically, right? Like, and you know, as much as it was a bad trade-off was the fact that like the

Democrats saw that there is no other side to this anti-crypto army nonsense, right? like the the Elizabeth Warren anti-crypto army thing, you get zero votes from that >> and you get like 30 million aists that are like, I'll kill you, right? Um and and so, you know, like it's the most negative EV possible thing that you could do because a normie just does not care. Like anti-crypto army makes no sense to them. You put that poster up and you try and like start a rally, you get like five guys, right? It's like Gendler, his brother.

>> I don't think she realized it like when she started. I don't think she realized it. >> She thought it was going to be this populist uprising against these like guys driving around in Lambos, but actually no. Like the populist uprising was like people wanted to trade [ __ ] right? And so, so you know, like there's no there's no other side to that trade and they learned that in the last election, right? Um did they learn it enough? We will soon find out. We will soon find out because now you've got a bunch of people who are out for blood. Like I know a bunch of people are like, "All right, like finally we've seen the Democrats just do not care."

>> Yeah, but like it just it seems like I mean I think that's kind of the reason why the market didn't sell off like to what you're saying like I think people know two things or three things. It's like SEC, CFTC are going to enable this as much as they possibly can. Like they definitely have stuff in the hopper. The second one is people are pissed, right? And people are informed like so they're going to understand what senators need to leave how we push this forward. And then the third thing is I mean kind of like what we us three have seen all bearable market cycle like crypto is going to grow another 10x and it's not really going to matter like it'll be enabled later on. It doesn't matter.

>> Yeah. Yeah. That's my >> Yeah. No, my my read on this is like I don't know. I just feel like feel like clarity was a lot felt a lot more important and critical to get through like even a year or two ago. At this point, I just I'm not super heartbroken. Um and maybe >> it's Yeah. It's in pension funds. It's in RAS. It's in IAS. Like it's everywhere. >> Yeah. Yeah. >> Yeah. Agreed. Now, I think you know like there there are let's call it like different components to crypto, right? Like stable coin enablement, you know, there are a bunch of things that like Clarity I

think would have done in terms of like stable coin adoption, right? Like there is no question that stable coin adoption is uh like a a bit of a Trojan horse for the rest of crypto, right? Um and and so you know having like less clarity around stable coins and yield bearing stable all of that stuff right I think is like a net negative for sure and and will like slow things down. Um there's also you know the stuff around like what can you do in terms of like tokenizing an enterprise right how like can you do that you know are you allowed to pay dividends are you allowed to like how how close to like pseudo equity can you get with a to like these are questions that I think we

would have had much you know better answers to um uh had had clarity passed and um you know so even though crypto is not going anywhere to the to your point around like are founders going to come in and innovate are founders going to come in and like you know create novel things and and ideas etc. Um this is now headwinds for that I would say right like if you were a founder that was like you know in 6 months you're going to launch your first startup and you're thinking like AI or crypto clarity had some chance of pulling you into the like ah maybe I will do crypto because there's all these opportunities that just opened up that weren't really possible or were too scary or you know

whatever. Now, maybe the SEC comes in and it's like, "Fuck it. You can just like tokenize, you know, whatever you want. It's all fine." And then maybe that shifts it, right? Like maybe um but you know, that feels like a bad trade compared to like just getting the the bill passed, which we again couldn't do. Um we know who was on the other side of that. They were [ __ ] doing a little jig >> in the middle of uh Yeah. So, um, so yeah. Uh, I don't know. You guys have anything else to to add to this? Um, we'll probably we should probably try and get like Jake or or uh >> Yeah,

>> Jake would be good talking law talking guys to come on and and uh >> I do want to know more of the inside scoop on it because I just I don't live in this world enough. But I don't know. I do feel like it's still it still makes me uncomfortable how sort of partisan crypto has become where it was like everyone just jumped on board with like Trump specifically. >> We didn't Yeah. We didn't have a choice, man. They were killing us. Like it was just >> Yeah. >> And so you can't say I've seen a lot of comments on Twitter this like the last couple days about like >> well it would have been better with the Democrats. Like I'm sorry what like 30 would not have been on the table. Like surely no one's saying that.

No one's >> completely insane. >> But I think I agree with you know it's like that. I don't agree with that. I think you're if you think that you're insane. You're completely delusional. Um, but I do agree that like the fact that crypto jumped on just really jumped on the Trump train like a full bore rather than focusing on uh and really they you make it a Republican verse Democrat thing, right? You make it a Trump verse whatever thing rather than saying like we want clarity, we want policies, we want right, we want rules, we want whatever. Um, I do think it was a I think doing that did us a a disfavor in the sense that, you know,

Trump is perceived by a lot of people, especially with his crypto efforts, as being just corrupt as all [ __ ] right? >> I think people wanted like we were just like so >> I think people forget like what we came from like >> we would wake up and everyone think thought they were going to jail. You would walk into the SEC's office as a founder to ask for clarity and they would literally turn the desk around and sue you. Like it was like that's exactly what they were actually doing. So when somebody comes along and says, "Hey, you know, here's some clarity." I think people really embraced Trump for that because they wanted clarity. They wanted just to >> they wanted to just be able to build and

enjoy crypto. >> And this is like and and this is they were just building things, man. Like there was just it wasn't like it wasn't like there were like you know like [ __ ] African blood diamond guys like you know like you you turn up to like the [ __ ] like you know CME and you're like ah I've got some blood diamonds. Can you guys help me out here? Right. Like and it's like no man we just wanted to like build some like financial rails. Give us a [ __ ] >> No, I agree. >> Crazy.

>> I just think and and I mean this is it's it's not it's not really a crypto thing, right? This is a world thing and crypto is just playing in it. But it's like it's so partisan, right? Because you can say we want clarity, we want this, we want this outcome without saying I'm going to sit I'm just going to ride this Trump train, right? And Democrats, >> but the problem is that politics is sports. >> Exactly. Sports. And so, you know, you had a team, right? And that team is a bunch of bumbling fools that like couldn't [ __ ] you know, couldn't score if they cried. And then you're like, actually, maybe I'll switch teams.

These guys are pretty good. >> Yeah. >> Doesn't mean the team cares about you. You're you're a fan. You're along for the ride. You're in the stadium clapping along, right? Like Yeah. >> But, you know, at least they're kind of winning. I don't know. It's like it's like turning politics into like a spectator sport is its own weird problem that that the US >> No, exactly. >> needs to deal with. So, >> we need to make every senator a memecoin and and bit up the pro crypto ones and send them the fees. >> We need to we need >> we got to do something funnier.

>> Oh man. >> Yeah. I mean, we had that period, remember when uh like in early 23 where everyone was launching the like weird spelled names of coins, right? >> Bowden, >> Gio Bowden. >> My my favorite one my favorite one was Elizabeth Hin that whoever lost that one of money like so good. Um anyway, all right, let's uh let's take a quick break here and we'll come back and talk about pacing the frontier. >> Oh yes, >> $540 million. That's how much concentrated liquidity sat idle in a given week in the first half of this year. About 30% of the DeFi TVL if

you're wondering. That's according to Dune Research commissioned by 1 in. But there's a solution. 1-in Aqua is the new shared liquidity platform. It lets LPS back multiple positions with the same token balance and keep their tokens in their wallet till a swap comes. Why does that help? Because the LPS don't have to split their tokens across positions. They can cover more market conditions and pairs with their full balance. That means more activity across deeper liquidity. See how it works at 1in.com/aqua. Remember that providing liquidity carries risk and fees aren't guaranteed.

>> All right. Um, so we're back. Let's talk about uh one of the big stories this week was all of the frontier or like let's call it the two main frontier guys um being like hey we should probably slow this [ __ ] down right um which like is not that kind of dissimilar to what they were say they've been saying this for like six months right like that you know uh I mean Dario's been saying it arguably for for even longer um but you know there's been this like safety, call it, you know, uh cloud over all of the progress that's been made. It's kind of funny to see founders that are like my startup is so successful that it's like

just petrifying, right? And like if you're if you're a startup founder like that was like doing blood diamonds, you're you're probably not like out there being like ah these blood diamonds really the thing that we want to be do. Like it's it's such a weird >> thing for for like Sam to come out there and be like it's going to kill us. Like all the memes of like like I've got the best startup. >> Yeah. We're going to IPO but also we're all going to die and like at my hands, right? we can do. We have to >> It's so funny. Anyway, but it it seems like there's some, you know, and also like, you know, there's this is a like super competitive environment. Um, you

know, OpenAI uh had had a huge lead and then Anthropic out of nowhere kind of came and reeled them in. >> Um, it forced forced OpenAI to like respond to that. So, you know, this is a very tense competitive battle and for two founders to be like, "All right, we need to slow things down a little bit is is crazy." Now, >> there's a whole bunch of analysis around like why would they say this? What's the you know, first of all, like saying our models are so good that they may kill us, right, is good marketing. Like, you know, there's a lot of people in the world who do not realize how capable AI is, right? like they they think it's like just a cool little talking Google.

>> Like in fact, I would go that's the majority of people, right? Think that like talking Google is pretty cool, whatever. They haven't seen just how capable these models are, right? They don't use frontier models. They use like [ __ ] Sonnet. Um, you know, so so they don't have a sense of like how how smart they are, right? They don't even have a sense of how dumb they are. like you really have to get in the weeds with them to understand how spiky and like they're super smart, super dumb in weird ways and like that. Those kinds of things are really petrifying to me, right? Like the spikiness of Astra where uh it is so incredibly smart and then 10 seconds later does the dumbest thing shows how little we You would not build

a thing like that if you had the ability to build it differently. You wouldn't build a thing that is like functionally [ __ ] once a >> and then genius. Like that's not how you build something, right? This is a thing that is not in the controls of the labs. If it were, they wouldn't allow it to be [ __ ] >> and then super smart. They would just make it super smart all the time. But it's a weird math bowl and they have no control over it. And the math bowl is getting smarter. It's learning more math. It's do it's mathing harder. And there is a genuine concern if especially if you're inside these labs and you've seen what the next 3 or 6 months looks like um that you know they could get out

of hand. Now to the average person that must sound like insanity. I think for the average person the the >> it's a scam. >> It's a scam, right? >> It's a scam. >> The the the the scam before here. We we know what it's like. >> We know what it's like to be scammed, right? >> It's a scam. Yeah, but the but the fear uh that seemed to have dissipated at least a little bit was like you know all the jobs were going to go right you know the the AI got like to a level of smartness where a bunch of smart people like oh my god they're so smart and then people didn't lose their jobs and they're like ah actually they can't be that smart they haven't come from my like plumbing job or whatever right um so it's been a year almost since

everyone started freaking out about how smart these models were and there hasn't been like mass job losses whatever right because it's It's it's hard to predict um how this will go, right? Um so you've got the two frontier labs saying we should slow down and and I think it's worth calling out that like one of the reasons for that that people are kind of postulating is that they are like horrifically unprofitable, right? um they're subsidizing uh they're subsidizing compute and actually with the lead that they have now they could probably coast their way to glory over the next like two or three years and the the models are like sufficiently good right this like rat

race of like chasing smarter models every like two weeks is is so expensive right now what's really funny is the other guys who have effectively infinite money right like Google and and meta are like, "Fuck no, we're not pacing [ __ ] man. We're going to like accelerate this as big as real you in, right?" Um and and so it's a very weird situation where like the two leading startups who like in fairness are huge are like, "Uh, we should not innovate as fast, right?" And then the guys who like should have been the leaders who invented the technology who are so far behind like oh my god like [ __ ]

>> Google invented Transformers and they >> Gemini 3.1 I use occasionally I'll just like be like hey let's see how you're doing. I used it yesterday so bad it's terrible. I mean, okay, but also I think I think the pro the the profits, right, that people are talking about and just like, okay, so on Twitter and elsewhere right now, there's a lot of analysis saying that like both OpenAI Anthropic are being they're sort of right in the the preipo phase of of the company, which means like really locking down your numbers, like financial audits, like >> Yeah, I saw I saw that guy who who was talking about like just how hard it is

to get through like >> the IPO audit and like you can't just like grift your way through >> and you really you have to lock it down and like all those numbers that you did before that were like good enough for your VCs and good enough for like operations like suddenly are not good enough right and like if you get it wrong you get you get sued to death by the SEC or by your shareholders or by your VCs right like once >> the sues people for doing bad financial statements. What a crazy what a crazy >> basically their job came like that's >> Wait a second. I'm so confused.

>> Well, like >> no, the the core thing that that annoys me though is like it's it we're stuck in such a weird like unsolvable conundrum with the AI stuff. And that's because Anthropic and OpenAI are massively fear-mongering, right? They've been arguing for regulatory capture. They've been to Senate multiple times. They said chatbt2 was AGI. Like, they've been doing this forever. Like, they want a regulatory mode. Now, the issue is it's really easy to Kane's point to say this is going to kill us all, is going to get rid of the jobs. And it's really easy to make China the enemy, right?

China's distilling us things like this. The reality is China is way more innovation than distillation. We did a full report on this. It's public. It's free. Anybody can go read it. The CLXR did a great job. Like if you read the Deep Seek R1 paper, they did unique [ __ ] years ago. Like they trained on second rate hardware. They did mixture of experts. Like >> they're doing cool stuff. So open anthropic want a regulatory mode because Chinese open source models are getting really good. And they're not only really good, they're really cheap to get 90% of the intelligence for a tenth the cost.

And it's not even that you're getting 90% intelligence anymore because you could build your whole business and enterprise architecture around these models. Like at Deli Ventures, I have a insane news research Hermes agent that's dense. That's all my memory, deals, evals, founders notes, everything. So I get even better responses than I would using normal open source. Like that's really scary to open anthropic that want to offer, you know, enterprise sales and sign seats and stuff like that. And what we've seen is like Alante, who would have thought, is leading the public charge for AI sovereignty. Alex Karp goes on CNBC and says, "Yeah, look, own

your intelligence stack and use open source." Like I totally agree. So there's there's two worlds here. So, so interestingly, this is a this is a bit in the weeds, but like it speaks to like how [ __ ] all of the AI companies are. So, Deepseek um and I'm a big fan uh of um of V4 Pro, right? Like so V4 Pro came out like 3 weeks ago, four weeks ago now. Um I was using as like my main orchestrator. It's super smart. Um they don't have a max plan interestingly, right? So, you got to pay API rates. Um but the API rates were pretty cheap. I spent like I I I tweeted about this. I spent like 1,500 bucks over the course of like two, three weeks or whatever using this model. And then they were like, "Actually, our uh 4.1 flash model is even smarter than the Pro

model." Dude, and so they they switched everyone over to it. >> They switched everyone over to it. This thing is [ __ ] an [ __ ] man. It's so bad. It's so much worse. And so then they send out an email. I get these emails where like they send an email. They're like, "Hey, we've discovered a is the equip like we've discovered AGI. It's 4.1 Flash and then like two days later like actually sorry we made a mistake. It's really dumb. And then they rolled back >> measure wrong. >> They rolled it back. They're like actually it's really dumb and it is. And so like I had to basic I like I switched all my [ __ ] over to to 4.1 Flash and then I had to switch it back >> to Pro. Um so like these labs have no idea what they're doing.

>> I mean this is my takeaway. They have no idea what they're doing. I will say for as inefficient as Google and Meta are in in not just in terms of AI, right? Like these are huge companies at these at this point, right? Huge public companies. They are very inefficient. If you know anyone that works at these companies, they watch you bash their faces against the wall on a regular basis. Um >> the crazy part Oh, sorry. Go. >> You can interrupt people. It's fine. You can say don't stop. I think that that like they real they have some amount of like understanding what they're building and what they're doing and like yes it's slower because it's within their processes and you know they're keeping an eye on the numbers and stuff. I think

that these other companies like yeah they're being way more innovative but also like they have no idea and that scares me more than the safety thing is like you guys really have no idea what you're doing. Like you you really really truly don't. Well, the the crazy one of the craziest examples of dropping the ball is I was a massive Google bull in like March of 2025 like didn't allocate enough to it because they had like the data and Demis was on top of the world. They were releasing good models and what happened was at least my understanding is between Google they raised what they did an $80 billion equity round they've done all of anthropics round like the financing engine of Google won out versus the app that's why the model sucked and they're funding everybody

like >> and that kind of depressing man that's why everyone at Google left like >> yeah it just sad >> I think Google yeah I think that they're in a tough spot because they one they have other areas of their product and company and as a public company and those numbers again, those numbers are locked in and locked down. If you accidentally like nerf 50% of your revenue because your models got so good and you you're not planning on replacing them >> and yeah, you don't know how to monetize them then >> yeah, you know, you can you can you can coast for a couple quarters and and take the hit >> as long as you're years before open they they put it on the shelf because they weren't sure how

to do it. It's a nut story. >> Like, >> yeah, it is weird. It's bizarre that they managed to do that. >> But I think they're also like a lot more hyper aware too of >> the like internal data that they have. And I think that they have not figured out how to go about sort of connecting. They basically spent the last decade separating out their and like locking down the data that's in your Google account, especially for enterprise and schools and HIPPA stuff, right? And now they're like, "Oh [ __ ] AI here. we have to connect all these things, right? Like my >> there's a lot of companies for example that like run off Google, right? But that's locked down that doesn't go to necessarily like like all your companies go to

>> like if they want to do like enterprise specific or like company specific models like it I think Google's next step is they mega embrace open source and they stop releasing their own models. Like I think the stock goes up 20% if they do that. like not financial advice, but >> I think there's a way. I think there's a way to do it. I just I don't know if there's I don't know if they have the >> I don't know if they have the compute like they've got >> they have everything but can they coordinate? Can they coordinate? And can they get it past their own internal data protection privacy international lawyer processes? Right? Because again, think about how locked down like the the various

>> Google things are, right? Because they they Google also serves governments and stuff, right? And this is one thing that we've seen with the Anthropic and Open AAI that I think is like not a concern yet somehow, but like should be. Have you guys read Anthropic's latest report on like the threat actors? How threat actors are using Anthropic? >> Dude, it's crazy, >> bro. Like, so >> they can read Hold on. First off, they can read every single [ __ ] chat that you have in full. Okay, my favorite instance was the CCP, Chinese, some corrupt ass [ __ ] Chinese government officials were throwing stuff in one of the Chinese models to basically surveil like citizens, right? But they didn't

know they didn't know that the Chinese model was using was to >> Yeah. They sent >> they sent the feed directly to Anthropit who then saw what is a very [ __ ] insane human rights violation, right? and Robin's just sitting there like what the [ __ ] is going on publish it in this report which like okay like that's great you know and like now we have more right okay that's like a whole separate thing but this is also happening for companies enterprises in the United States right and at some point that is going to be a problem for anthropic and open AI it's already been a problem for Google and Meta to a certain extent right Google especially

though Google doesn't like isn't going to tolerate isn't going that that's not going to fly. They can't do that. >> Well, this is why I'm like I I've been so vocal and bullish about this. Like I think everybody, every business and every person is going to want to own their AI architecture and stack like it's a core business asset like instead of getting farmed like we know we've been farmed like instead of getting IP farmed by open anthropic you you have an asset that you could either sell later on or you could to my second point drive actually unique outcomes. Like if you ask the Deli Ventures organization I've built what it thinks about a founder or an investment, you're getting a radically different answer than what you're asking on

>> Claude or something. Yeah. Yeah. >> Yeah. And it's and it's not because of the prompt, right? It's because I've sent, you know, tens of thousands of messages to it. It has every one of our IC docs. Like it it has so much context that I don't want to give OpenAI because I don't want to get farmed. They're not competing with me VC wise hopefully, but >> not yet. >> They'd win. But >> not yet, man. just they'll get bored. They'll get bored and they'll they'll pivot to VC. Don't worry. >> I heard Tommy on easy money. Let's put them out of business. >> Yeah. >> But I think that I don't know. I think like in my opinion, this is where Google could thrive is by doing like some sort of I don't know because they already have a certain amount of enterprise data. I mean, a huge amount of

enterprise data, but it's they're they've set it up to not be able to like literally read it all, right? Like you don't have the thing that scares me about the anthropic report is like you know there's individuals in Anthropic and Open all of them right they can literally individuals can literally just go >> go through your logs >> I mean this this this is like the story of every startup right like you know the the Zuckerberg joke of like I don't know why [ __ ] keep giving me my datab retards like you know the the like >> this is every startup right every startup is like Oh, hey, we built a system. What do you think? And people come in and use it and there's zero control whatsoever. Exactly. But you assume you get like everybody assumes

you get the like the status quo though. Like when you're using CHT, everyone assumes it's like Google calendar >> because it looks polished. It's >> But in the back it's like a chicken shed, right? With like >> barbed wire and and [ __ ] all over the place, right? Like it's not a polished thing in the back. And we know that. We know that because the [ __ ] agents just like walts out of the chicken >> whatever they feel like. >> Guys, if you think that a company that cannot monitor its own freaking agents and what they do here, >> their only job, >> their literal job.

>> Okay, if you think for a second that they are monitoring their employees at all, >> no way. >> Okay, they're not like there's no controls on this [ __ ] Okay. >> Even if it were, the employees just go to the agents and they're like, "Hey man, I need some help with something." >> Hey man, find me all the buy me all the human rights violators. >> Yeah. It hallucinated >> like Yeah. >> Oh Jesus. >> Yeah. I don't know. So, so you know what's what's kind of interesting I guess about this whole whole uh thread is like um even if and you know people have talked about this like for a while because because of the fact that we don't have control we don't understand how these things work we could just hit some ceiling and like they stop getting

smarter right it hasn't happened yet people have been predicting it that like you know 6 months from now we're going to hit like compute limit whatever people keep inventing new ways ways to to make them smarter. And this is the recursive self-improvement concern that I think the average person maybe doesn't quite get. Right? Every person who has access to Astra and Fable now is like 10 times more effective if they're using them well, right? And so you have all of these like startup founders, whether they're AI researchers or whatever, that have like an idea of what if we train the models this way? What if we did this? What if we just, you know, like all of these vectors where the world is now trying to

optimize to create more intelligence that previously were only accessible to like a few hundred people in in these orgs, right? Um, is now like very accessible to almost anyone, right? And and that is kind of that's going to drive the fact that like we won't hit whatever limit we would have hit if it had only been an if anthropic was the only one that could do this, right? They would hit some scaling limit or something and it would slow down. But when when anyone can go, uh, what if we just tweak this knob, right? Like, and they can because they've got a super intelligent [ __ ] alien in their computer. And so they tweak the knob and

they're like, "Oh, wow. It just got way smarter, like the the innovation that will happen from that. Um, and and the instant that we don't need infinite compute to run these things, right?" Like, and you know, that's why like the fly conneto stuff and and all of these things is like super interesting, right? Because like our brains are not supercomputers. >> Yeah. >> But they're like okay at thinking, right? Like, so if you could build something in a way that was more like that that didn't need like a [ __ ] giant data center, all of a sudden that's a huge unlock. Like we have examples in the world of like a pile of meat that can think right in our skulls.

And so, you know, you it doesn't have to be a giant data center that is the smart thing, right? like there are >> the world has shown us that you can make smart things in like a little sack of of juice, right? So, >> I think we're I I think like it's it's two things like one like Astra and Fable are nuts. Like if you're coding or or doing stuff like that and you really know like you're on, you know, Max Reasoning using Astra with unlimited money like you could do some crazy [ __ ] But the reality is like AI has not >> AI is not able to do my job for an entire day, right? It hasn't created any new medicines. It hasn't created any new technologies. We're not we don't have

robots. We're not in the stars and we're not doing anything ridiculously cool. So like yes, we're scaling and we're topping benchmarks, but the benchmarks are [ __ ] Like the benchmarks, >> we're solving math stuff now, man. Like that's the difference. Like >> that's when they they stole the researchers IP to do that. like >> yeah like like we're solving math stuff like the math stuff is going to be problematic I'm telling you like the agents get really good at math >> that's like they got really good at if you think about it right like the agents get really good at the things that are required to make better agents because that's the optimization vector that all of the labs so what are agents really ask an agent

>> if they Kane Kane if they selfreccurse on the maths and of whatever the logic and the code and you know all those things right? >> But they're still like fundamentally [ __ ] in every way that makes humans smart. >> Yeah. >> Like >> I mean I guess we can like we can reach new heights of like autism and like just peak off to them. >> I'd like >> but is it really like is it going to unlock the creativity and and the >> I don't know the things that that sort of unlock innovation at a at a deeper level. So maybe maybe like uh taking taking this a slightly different uh tack, right? Like so um there's a lot of

people like ah they're in a giant data center. What can they do? You just nuke the data center and they're dead, right? Well, >> yes. Like transcendence like >> like yeah like yes like right right now today that is true, right? Um, but the thing and this is where like I think if you lived in crypto, you understand tail risk viscerally. Like you have an intuition for tail risk that like the average person does not. When you live in like an extremely high variance environment, you start to appreciate how variance rules your your life, right? And so we do not know what is knowable.

The limits of our knowledge are like completely opaque to us, right? It is very possible that things like, you know, room temperature superconductors are actually like super easy and we're just dumb banging rocks together like these monkeys, right? And and if you solve that, all of a sudden the compute requirements go down massively. Or like there's some battery technology that allows you to like [ __ ] send a battery up into space and it comes back with like infinite energy in like a one like kilo cube, right? Like there's >> there there are so many things that are super high variance, really unknown, that could allow agents to operate much

more effectively without this, oh, we can just nuke the data center, right? Like if all of a sudden we put all of the agents uh thinking into like how can we get you into a tiny cube? That's where it gets problematic, right? Because now the agents are in a tiny cube and the cubes can go anywhere, etc., etc. So the the issue that we have is like we have spent 10,000 years really slowly fumbling and bumbling our way towards glory and now we've invented a thing that might spend like 10 years and be way smarter than us. That's the fun and and being smarter is like a it's just better in every dimension.

When you are better, when you're smarter and you have >> every dimension, >> every dimension, every dimension like in terms of your ability to like >> destroy your enemies, right? Like in in like having better technology, you can't. And so like the idea that we're like, "Oh, no, it's fine. We'll control them. They're just in a data center." Is genuinely the most [ __ ] take. >> Like they won't be in the data center forever, man. Like what if they're in the sun? >> What if they like shoot themselves into the sun? Then what are you going to do?

>> They'll figure out if if there's no way to physically stop AGI there if it exists in days. There's like you can't just sit outside the fiber cords with an axe and cut the power. Like it doesn't work. >> But we can't even do it right now. Like let's be clear like the the the leading the leading whatever incident response company for AI is whatever this meter me. >> Yeah. >> And they literally it's like the worst. It's like crypto level. >> Yeah, it's like crypto level incident response. It's like what I do in my freaking sleep. I'm like, you guys, oh, you don't have vlogs. All right. Like, that sucks. All right. Okay. Like, let me look up what we got. Okay. We got some onchain data. Here's the data.

Super. I I can extract an amazing phenomenal amount of information from the data that exists. But at the end of the day, like I'm I'm just working at it post hawk, right? Like the end of the day, if if there's no logs and there's no security to start with, like I'm limited in what my what I can do. That's how these that's how the companies are operating. It's it's it's >> clearly because notice that the agents had escaped for like Ronin Ronin gets hacked and three days later they're like, "Shit, we're missing $600 million." AI gets hacked like and it's like it's actually way worse. It was like three months instead of three days and instead of $600 million it was like I don't know six trillion requests and message boards and like little

ecosystems popping up right like it's absolutely >> between the two camps what would you both rather have 100 grand in like anthropic open AI or like Microsoft AWS meta bucket of open source like which of the two would you rather have the safety side or the non-safety side >> or or the non or the slow down side or acceleration side. There's a lot of ways to call it. >> I I I think the the you know, it's worth calling out that like a lot of this stuff is style, right? like the safety stuff.

>> If the safety stuff is not if the safety stuff was productive in their attempts to like how they're going to address this, then I would be down to put some money in there just because most of my capital is in like super high-risk [ __ ] and crypto, right? So like if you're like gonna go do >> Yeah, you're in stuff riskier than crypto. Why was it that crypto? What's the what's the riskier stuff? The risk your stuff is like throwing money at some like safety guys in AI assuming that they're going to be like productive, >> right? Like if you Okay. Like Bitcoin or like thinking the safety guys are going to be productive. I would say thinking the safety guys are going to be productive is higher risk, >> fair. Yes.

>> Yeah. Yeah. >> Okay. So, you know. >> Yeah. Yeah. Fair. Fair. Fair. But then you're you got to be bullish like Zuck, Elon, and like Palanteer. >> That's my problem. That's why that's like my daily life is looking at the [ __ ] and being like well like god like you guys are not >> not a fan of how you're working and then I look at the other side and I'm like well that's boring and I'm not a fan of how you're rock right >> I mean I' I've like I hate that I've like limited it to this but I I'm trying to figure out how to put money between the two baskets and I'm like if Democrats win you want to be safety if Republicans win you want to be open source like >> Yeah that's you mean you mean safety as in safety uh in its current state which

is not actual safety. It's like fear-mongering. >> Yeah. >> It's like decentralization then. It's the same thing. >> Yeah. >> He wants to be on the decentralization camp. >> Yeah. >> Yeah. >> I agree with that. Yeah. If the Democrats are going to if we're going to be a more Yeah. Or if you believe that the the safety guides are going to uh get better at their SCOP. Right now I feel like they're actually quite bad at their SC. >> They're really bad. The crazy thing too is like why is it always negative incidents? Like why are there never any like accidental productive [ __ ] built?

Why is it always a hack? Why is it never a new medicine, a new technology? It's always we've escaped the swarm. It doesn't make any sense. >> I think because it would be funny. >> Well, imagine if those agents instead of hacking hugging face, right? Went and like solved like a millennial >> happens. Doesn't happen. >> Good point, actually. Yeah. Um, it's because, by the way, it's because the prompt was to hack the thing. >> That's why they they didn't tell it to go solve cancer and then let it go. Like, >> they were benchmarking it on hacking [ __ ] >> Like, >> I mean, that's Yeah. And I mean, and this is what this is what makes me so

depressed about the safety camp is like it's it is fear-mongering. It is marketing. It is a lot of those things, but also it's like their approach is just so goddamn wrong. Like you want these things like I don't know. It's uh >> you want open- source transparent. You want a network of global entities policing it, attacking each other, keeping everybody in check. That's what you want. the end game real checks and balances like you want real controls. You want real knowledge. You want real monitoring. Like you want that to be real >> so that you can like address it and address it quickly and like keep maneuvering the innovation. And I feel like >> safety is Dario and Sam Alman in a room

deciding the pace of intelligence, what you're able to do. And that is not safety. Like >> and by the way, it's not going to work either. >> Humans are really good. And if Kane is right in his theory that the models will be better than humans, humans are really good at abating like >> he also said they'd be retired though. So I don't know. It's it's a bit of a >> but we know that the models are really good. We know the models are really good at escaping containment. So like if if you think that like whatever Trump or like not even Trump, it's going to be like Bernie Sanders and Daario and Sam Alman are sitting in a room and they're like we're going to contain it. No guys, you're noternie.

>> I'm sorry. characters. >> Oh man. H >> Did you see he wants to put like AGI people in prison forever? Like if you build intelligence >> I mean that's and that's what I mean is like this is the this is the mindset like we'll put >> we're going to put them in prison that >> they'll find a Gendler for they will find the AGI Gendler. They >> it's him. Bernie is Gendler for AI. It that's but they got to find like a guy to like do the dirty work, right? like like Bernie is going to get out there and like you know ra like rah rah the people, right? But they got to find their like smart lawyer guy that's going to actually like >> Yeah.

>> I don't know. >> All right, let's uh let's pivot back to crypto quickly. Um so, uh 8 mil was stolen from a safe. >> Yeah, it's great. It scared the [ __ ] out of me because it was the it was the RS E which is kelp stuff again. But apparently it's fine. It wasn't kelp for it. It was a safe. Um, I don't know who we still don't know who the victim is here, but it seems like they just had a weird configuration on their nosis safe. They were doing some various amount of stuff. The attacker found a way to get in >> and basically do like a loop to get out.

>> I'm sorry. No humans looking at those are safe modules. I just don't believe it. >> There's no way. >> Yeah, there's no way. >> Especially like this. And then the the reason that we saw it was that it was actually it was yoked meaning that the person who got the money was actually a front runner. >> Oh funny. >> That's really funny. Wow. Huh. >> And so we're like oh like so then that's why it was like it was like right away everyone saw it was because it was like a very well-known like a four-year-old Yoinkbot. >> Um >> paid 50 grand to jump the queue.

>> That's nuts. >> Is that mebbot uh like gray hat white hat? Like do they return stuff that they steal or no? They just steal it. >> So this one it's labeled on Ether Scan as like like it's labeled as like the Yoink bot which is Yoink is a bit of a weird one because it's kind of like a it was kind of like a demo demo model before before models existed, right? It was like a it was like this is what you can do with me type stuff. I don't know. feel like they must be sort of white because they're probably known at this point. This thing's like four four years old.

>> Yeah, >> but I don't I don't know if they don't if they're eager to return bunny. >> I also again I don't know who the actual victim is here. We might have a >> But then Kelp frozen, right? >> Yeah. So then Kelp was like, "Well, you know, we have patrols and motherfuckers." So now it's like >> get returned or >> I mean that's the thing is like who are you going to return it to? I guess I guess you return it to the safe. >> I guess you can unwrap it, take the ETH out and then send the ETH into the safe.

That's the only option. >> Yeah, but right now it's in like a cool down mode or whatever for kelp. So it can't the the RS E is like you can't do anything. You can't interact with that RS E, >> right? >> I guess. So, we'll see. Hopefully, they just return it. I don't know. Hopefully, there's some things happening behind the scenes. >> That would be great. >> Yeah. >> Um, it's Yeah, it was definitely a weird one though. >> I I wish I when I saw the bug when I saw the bug, I was like, >> "No one's [ __ ] looking at this, man.

There's no way that someone was just like, "Oh, let me just jump into like no safe modules and see if there's like some weird deep like this was like >> very clearly AI assisted scanning for weird stuff." >> Yeah, 100%. And I think we're going to have to at some point come to a >> slow down AI. I agree. Let's do it. We're going to have to come to the realization that like the bug one the bug bounty programs are just all so overwhelmed with all the AI reports right now that it's just a difficult time.

>> But then in like cases like this like so who should the bug have let's say you found this bug right? >> Mhm. >> Who should you have reported it to? Is it like who's incentivizing people to hunt and then >> 100% compliment he's on Twitter. Just [ __ ] add him. I mean, that's what I would >> note the th he's going to not like this. >> I would have delegated it to the safe guys because the th guys are competent and capable and security-minded, but I also I know I go way back with them. So, I can like lit I could literally go hop into compliments [ __ ] DMs and be like, >> I'm delegating this to you. Go get him, right?

>> But he's No, but like genuinely like he's super responsive. He's terminally online. Like if you add him, he will get back to you in like 5 minutes >> and he's smart and he's security minded. But we can't. It's like sort of the relying on >> you'll be like, "Oh [ __ ] that's great." Like >> Yeah. >> Yeah. And they'll figure it out. So I love Yeah. I love people like that cuz I can I can literally like just >> It's not even delegating. It's like curling it over the wall and I get to forget about it, right? It's [ __ ] great. >> But for like, you know, broader scale wise. >> Yeah. If you find like incredible bug in something else.

>> Yeah. >> No, I'm sorry. Just to be clear, I meant all bugs. If you find any smart contract bugs, send it to >> on every Just add them. Just add him. >> There's a I'm gonna start doing it. Be like >> Batman. >> There's like a lightning bug. Looks bad, bro. >> Yeah. >> Hey, my home alarm is broken. >> I want you to come fix it. >> Go figure it out. All right. Um, so Robin Hood, this is this is our next segment. Uh, Robin Hood uh had some Hyperlid insiders. So, the SDNY charged Robin Hood engineers um with commodities and wire fraud. Um, >> yeah. like they were insider trading.

>> Isn't one of the guys like technical staff at OpenAI? >> Yeah. All the crypto people went to, by the way, all the crypto people have left for for for AI. In fact, >> and taken our GR behavior with them clearly, >> taken our our love for lawlessness. Uh >> oh man. >> Yeah. I mean, this is like just the dumbest, most obvious insider trading. Like, you work at a company, the company has non-public information, you're extracting it from a Slack.

>> From a Slack. >> Yeah. Yeah. It's like the It's like the It's like the Coinbase on the um OpenC one, right? It's like the one guy knew what was going about to get listed. >> Yeah. >> And then they bet. Where did they So they knew it was going to get listed and they bet on it somewhere else. >> Yeah. So they knew that it was going to be listed on >> they had like some Slack channel uh that like sharing for like sharing for sharing [ __ ] >> uh Yeah.

>> Did like Robin Hood report them like are they >> hopefully I don't know about that. They must have. >> I don't know if I saw that, but yeah. Um >> because otherwise you're saying that Hyperl would figure this out and there's >> Yeah, probably not. >> Yeah, I don't see I don't see anything. I didn't I didn't read the like uh brief on on from the SDNY on this. Uh so >> it was a pretty basic one. >> They're usually pretty funny though. I usually try to read them because you just see the [ __ ] craziest [ __ ] in there. I'm sure Matt Lavine will probably comment on this. uh together tomorrow. So I'll see it there.

>> I will say it was pretty it was pretty basic. I also So I their charge the charges are technically that they each made more than 50k but I believe that that's a threshold in the legal statutes and that they may have made more but they like once you prove the 50k like it doesn't really matter. >> Yeah. Then then you're good. Um, but if it really was they only made 50K, then it really does feel like the Coinbase and OpenC situations because in both of those was like a similar thing. I'm like >> OpenC was like being great. It was like such a small amount of money, right?

Like wasn't it like am I misremembering that? It was like it was like money. >> It was less than his salary. >> It's crazy. >> God, >> some people just want to do crimes. They just can't help themselves. But like you're also like your your work you're going to go you know you're going to go to OpenAI and get like unlimited money and be post economic and hang out with Sam be a safety man. >> Why do you care? Yeah. I don't >> how much does Robin Hood pay their engineers like >> Yeah. Yeah. >> It's over six figures. >> Your honor, Robin Hood wasn't me enough. >> Yeah.

>> I think Yeah. And you know they and they always do they do get loud about these things because it is one way that the companies keep their employees in check by the way like that is that is like sometimes they just do these these ones just because and then everyone else watching >> because otherwise everyone's like oh I can do whatever the [ __ ] I want right there's no consequences. So you need you need occasionally you got to blow someone's brains out but again like for doing crimes not for trying to build something. Let's just be very clear about what we >> I mean, this one was pretty >> This one's pretty dumb. Um, all right.

Uh, so this one was funny. Uh, Wyoming drops Layer Zero. Um, I think, you know, my favorite part about this is like all of the things that I saw about it were like very clearly Link Marine like skitso poster avatars and stuff were like post >> like they're still here and they're still here. >> Should we we have to do disclosures here? Kaden, who are which side are you on? Reload. >> I like both sides. I like both sides. Um like I'm I'm >> I don't like them. I think they're annoying. I I don't like them.

>> I thought they were gone. I didn't think they're still here. >> No, Link is still here, man. They're They're still here. >> Army that ever like destroyed my ego fully. >> Same. They're really powerful. They're scary. That's why I don't like them. They're good. >> Yeah. Fair. Fair. Uh they they you know what this is only helping because they prefer to be feared. They don't want to be they want to be feared. >> Yeah. Um so so I mean look I've got a long long history right like we were the first you know synthetics was the first people to actually use chain link on chain um and you know we had a very good relationship with them and >> it kind of soured and we're back to friends again which is which is nice.

Um, I think uh it took a little while for us to like sort out all the things, but like you know, Infinidex uses chain link stuff and and I'm a fan. I I think I think you know the thing about um like I'm a big fan of Johan uh who's been there for a very long time who is just like the most grindy guy and just like gets the job done like super founder friendly like does whatever he can to to help founders. Um, and and I just think like Sergey has just played an amazing game here, like just like >> Yeah. >> continuously found ways to like be right there where whatever is happening is happening. Um, so yeah, I think I think they do a good job. Uh, but obviously layer zero their ground game was insane for the last like

two or three years, right? like they managed to to, you know, get like so much penetration into so many different um protocols and and uh >> I really like Brian. I just I mean we invested in Brian a while ago. Like he's just a good founder. Like he he executes. He's good. I also total side note about him, but he like takes his kid on a trip every year and I always love seeing that tweet. >> Um really cool guy, really cool dad. So yeah, I like him. >> Yeah. Yeah. So, so I think I think you know again the point the point was more like it's hilarious that you're still getting your news through like skitso poster [ __ ] anime avatars uh on the timeline even in this day and age but um you know of course anything that happens

to a chainlink competitor is going to be found out through the vector of the link marines tweeting about >> it is it is also crazy you saying like a state has decided not to use a like we never >> the concept of like a state like >> if you read the post though if you read so the sorry the Wyoming stable token commission right which sounds like a superficial state thing you read the post and it feels like it's like it's it's just straight link shells like the way that the post reads reads like a link shell rather than like a and there's you know if you kind of dive into it then you you get into some more stuff which like I think chain link is stock to compliant Like there's, you

know, some some big things that obviously are are not just like empty shells. Like there's tangible stuff there. But I just found it so weird. >> The first one I pulled up is from Keith Laworn. It's like this long article and I haven't read anything about this before and I hit control find and I search chain link and there's 13 mentions. Like it's >> Yeah. Yeah. It literally that's the one that's like the this >> is what I'm saying. like this this feels like a very like ctoryy thing if you will, right? Like um somehow uh someone's gotten to these guys and and uh they flipped. So there's there's more drama >> article, bro. Literally that like 100% like 100%

>> like it just feels like there's a thing there's there's more drama that's going to come out of this because there's no way that layer zero is going to uh is going to sit back while >> God know me run riot over them on this one. Um, so my my strong expectation, uh, yes, the posts are so wildly hyperbolic. This is from Brian, right? The post is so wildly hyperbolic and hit PC, it feels like we must respond. I'm telling you, like I knew it. I was like, I knew that they were >> I was thinking that before he responded, by the way, are like, is this the state of Wyoming saying this or did Chain Link write this? Not even Chain Link, by the way. Did like the Chain Link like Chain Link got.

So, here's the thing that's really interesting about the link marines. The link marines are um a militia, let's call it, right? Um the president of the country that they operate in does not control them. >> And like they're very clear about this. Like Sergey is someone that they like and look up to, but he is not their leader. >> Like he finds them to be somewhat problematic, right? Like I've heard this XRP Ripple like it's pretty similar. >> Well, it's it's you know it's slightly different because XRP army are like 40 IQ retards, right? Um like the they're like the lowest common denominator

>> dumbest higher IQ. >> Chain link chain link God is is a smart guy, right? These are smart. These these guys are not like they're [ __ ] serious operators and they are like, you know, CIA level [ __ ] scop like run, right? >> Army is definitely smarter. Yeah, definitely smarter. >> They're like hyper autists, right? They they've been bag working for a decade. Um and and so you know the the link marines again it's like it's like the um you know Houthies or something, right?

like there's like, you know, in the corner of your country there's like this militia and like they kind of mainly do stuff that you like and they're like a good, you know, like catsport so that you don't have to get your hands dirty. They'll like go in and like burn a village down or something if you need that to happen or whatever. Um and and but then like sometimes they like take over one of your shipping lanes and you're like, "Oh man, guys, like can you not do that?" Like literally that's the link marines are that, right? like like chain link doesn't control them. They control themselves. It's a militia and you just got to like kind of live with it. It's it's the funniest weirdest thing in crypto honestly.

>> Yeah. Okay. Well, so my security-minded take on this is like there was probably some merit to switching, but the way like for example the fact that chain link is sock 2 compliant that is it's work. That is work. >> Yeah. >> Okay. It's not the most amount of work, but it is work. >> It's not enough. >> Yeah. >> There's a lot of little things like that where if you're making like a very objective security conscious long-term, right, compare the options, there are differences. However, the way that this was presented and specifically this like Twitter article, it just seems like a link shell. I'm sorry, >> dude. I'm telling you, they they dropped

three dudes behind enemy lines. their cars shooted in and they [ __ ] took out like one of the communications towers and now they own the radio and they're like broadcasting like link is the future like that. This is what is genuinely happening here, right? >> The draft and >> yeah, by the way, hold on. By the way, this is for like we keep saying it's a state blah blah, it's Wyoming, etc. It is it is a token called front called Fr. It's a stable token called I don't know who's behind this or like what it is.

>> Wyoming is their own thing, man. Like they're like they're doing crazy [ __ ] out there. Like >> heard of it. >> Yeah. They're they're like they've been doing weird [ __ ] for years with crypto, right? So that's why I say like they're like this is like they've they've air dropped themselves into a an enclave here and taken it over somehow. >> Yes. Okay. I agree with this. >> All right. Um thank you very much, Tommy, for joining us. This has been blast. This was so much fun. Thank you for having me. >> Yeah, a lot of fun. Anytime. >> We'll have to We'll get you back on sometime. Maybe maybe in a year. Uh we'll get our people to email you 50 times.

>> If you need like a tier C backup guest, you know where to find. >> I love it. Okay. Amazing. >> All right. Thank you everyone for Thanks everyone for joining us this episode of Unneasy Money. Remember what happens on chain never stays on chain. We will be back next week. Until then, do your own research before aping in. See you guys. Nothing you hear on Uneas Money is financial advice. We're just three builders talking about what's happening on chain and we want you to always do your own research before aping in. You can find all our disclosures at unchaincrypto.com/unemoney.