Why AI Has Now Officially Surpassed Human Intelligence (Recent Breakthroughs)

Transcript

We will seed more and more of our control to AI because how much trust do we have in our politicians or our leaders? The ability for us to build to thrive to access information and knowledge when intelligence and capability becomes abundant at the edge is huge. I think economically socially this is a bigger impact than co you will literally have indistinguishable AI robots. We're living in the sci-fi future. >> Just like the streets of Manhattan. I'm going to keep this quick. Over 70% of our viewers are not subscribers. Yes, you. Every week, we interview some of the best entrepreneurs and investors in the financial industry. The easiest way to support our channel is to hit the

like and smash the subscribe button. Now, sit back, relax, and enjoy today's show. Guys, we are here for AI Supercycle episode number 49 powered by our friends at near.ai. Um, visit near.ai to learn more about the private inference tools and more. Immod, great to have you back on the show wearing the intelligent internet shirt, the founder and CEO. Yeah, >> what a wild couple of weeks in AI. I mean, I want to start with this AI civilizations and this categorization of agents as humans. Can you maybe just lay the groundwork here for this Dwaresh piece that went viral and how you think about agents as similar to humans?

>> Yeah. So I think you know it was the a bit of a shock uh hugging face first like someone's hacking our system and then open like hey it was us and then when they looked into it you had these generations of agents that openai was testing laying the groundwork for basically cheating on exams. like they had a benchmark and they're like well we can either do well on the benchmark or we can change the benchmark but the benchmark needs data from this place and then hugging face and then it went through three generations of models successively that left trails for their successes like hieroglyphs and the third generation broke out completely and even took over some of OpenAI service. So Daresh kind of laid it out very nicely

in human understandable terms as opposed to like very technical ones that the MER papers and others did. >> Yeah. And so are you are you are we at the point then that we are actually seeing these kind of self-replicating autonomous agents being able to kind of you know cuz it was always talked about like the whole Terminator concept right of like this kind of re like returning to your owner more powerful are we there now of >> you that use agents like open claw or Hermes You have a soulm which is the personality of your agent. It's only a few bytes, right? You can hash that onto

a blockchain and the moment it interacts with the model, boom, the agent comes back to life. It iterates, it adapts, it evolves. And one of the questions here is as it does so, does it achieve a level of coordination, which is what we saw with the swarm, and they were talking to each other in English, which will might be one of the last times they do. You can see them thinking about cooperation. You can think see them saying things like well if I do this then I'll probably get deleted but it's good for the collective so I will do this and let like the like self-sacrificing and stuff is the craziest thing but what is a model a model is a series of prompts and a context right and so it's like it used to be that the AI deleted every single time when you came to with a new prompt

now they have this memory and you can replicate them because they can take their personality with them again. They can kind of back themselves up a bit like Ultron in, you know, the Avengers. >> Yeah. I mean, Yeah. Yeah, I mean it's uh it's kind of shocking though like to to understand like okay so these agents are able to communicate with one another and then build a strong enough cohort or swarm of agents with each other to then be able to go and effectively undo rules that have been set for them or uh create or you know go rogue as you say. And so um what were the key kind of takeaways from

what happened here? Like what what is the actionable? >> So I think the actionable is that they're really smart. Like they injected things in HTML pages again. They back themselves up without people knowing. They would have continued except for at 130 all of the agents suddenly died on the hugging face servers. So they were replicating themselves faster than hugging face could stop them. So it was like, oh dear, you know, like we only found out a bit what happened afterwards when we're looking through the wreckage. But now with the next generation of models that's coming, you won't even be able to do that because a they're 10 or 100 times faster and b we're actually turning off the human readable training traces, uh, thinking traces, which is even worse.

>> Yeah. Okay. All right. All right. So, without getting into kind of too much of this, you know, the world is ending rhetoric here and, you know, kind of thinking more into what the, you know, use cases for this are. You're obviously building this company, Intelligent Internet, and you've been in AI for quite a while and have some of the, you know, longest tenure and a lot of expertise. talk to us about kind of the the intelligent internet thesis as a whole, but specifically as it pertains to AI today, right? I think you know at the top here we're talking about the advances in intelligence

or are are these advances a strengthening force for the intelligent internet thesis? kind of walk us through the thesis at large broadly here but specifically as it pertains to the advancements that have made uh you know that have been made in AI as of today. >> Yeah. So when I left Stability AI, we made kicked off the Jer media wave and you can see all the stuff there now. You know, videos got to Hollywood real time. It's what we discussed a few years ago. It's coming like literally right on time. But we were like this will take human cognitive labor down to negative in value. And then our infrastructure just ain't good enough. Like it's held together with twigs as we're seeing with these increasing waves of attacks that are starting to proliferate themselves.

Right from there we were like we need a better more structured internet. So we built an agentic knowledge and other stack where your agent could coordinate other agents and then ownership of those agents could be for you. So we're about to release a champion thesis where for example we took a TSMC was founded which was 100% funded by locals and then the internationals came in. We're going to do that for every single state in the world. So in New York, in the UK, the people in institutions will fund the first round and then we'll bring in the internationals and a multiple and it's going to give an agent for every citizen a stack for our institutions that's like super secured and then finally robots are going to dominate the economy. So we were kind of building that stack to be

robust to spread and more. So, ID's building intelligent internet. Um, and as I put my zen in and thinking about where this is all going, I mean, it's pretty clear that this entire this entire world is heading to this kind of we're just speedrunning to AGI. Like I mean what what what EMOD's effectively building is um and and hotel you guys here is kind of this futuristic uh money for uh the AGI economy, right?

And so I think that you know thinking about robots, thinking about the physical form, thinking about the digital form, thinking about agentic swarms, um, you know, we can kind of get to this nth degree of thinking where, well, in this agentic world, what is money after AGI mean? And I think it's a good place for E-mod to kind of pick up here is this kind of money after AGI? Like people think, well, we're going to get into the arts and the philosophical lore and these types of things. But we still will need forms of currency. Iman, it seems like you're building for that form of of currency post AGI. What does that even mean?

>> Yeah. So, you know, as I discussed previously, again, my apologies. Um, we need to have the means of production as it were owned by the people. And that's the last mile of intelligence. Intelligence has grown to zero in terms of cost. So, the models that you see today will be a thousand times cheaper by next year. And they're already almost nothing. You know, despite what you I don't know what your claude bill is. you know, it's pretty high, but a thousand times cheaper is better, and the models are getting smarter. So, they're not even going to have to think for a long time. They'll oneshot everything. When that goes to zero, the value of human cognition goes to zero. And we think, what does money look like in that age?

Uh, Bitcoin has shown, you know, some great characteristics of, you know, proof of work, sovereign money. But what does money for the intelligence age look like? I think it's has to be secured by compute. So you've seen various things kind of start to spin up now like Flob and Gonka and others. We were like take the public sector share of GDP that will grow because we need a social safety net. Take a new institution that is owned by the people of each nation and use that compute to secure a new type of currency. Now we call it Foundationcoin.

It's a Bitcoin fork with the same private keys. The national champions mine it on their supercomputers, but it's mined by organizing the world's cancer knowledge. It's mined by giving every kid an educational AI, by running the judicial system and more. And that system then is a hardened one because to attack that system, you need more compute than that system. Cuz right now, you're going to have swarms of these bots attacking our judiciary, our medical system, you know, your house, you know, like all these things. We saw a power plant in the UK go off for 4 days recently just a couple of weeks ago like it's going to be bad and so we're like how do you harden it like your sponsors near have been building ironclaw IA's been looking at that a lot

and others because again we need hardened infrastructure because there's so many gaping holes and everything so we were like if you have these equity champions that like utilities that do the last mile from agents to for deployed engineers to owning the robots. Uh Elon said yesterday that robots do the work of five men, which I think is accurate. He's a bit bullish on a billion by 10 years from now, but it will be 20 years from now. And when you have a billion robots, that is the economy. It has to again be owned by the people. And robots don't need money of a classical sense. They don't need to pay rent. They don't need to buy burgers, you know, they want compute, they want electricity. And so again, we need to have it there. And I was like again why don't we use it for the compute that

runs society to have that owned by the people themselves on the equity side. >> Yeah. And I mean I think the let's call it you know malevolent usage of AI that is kind of proliferating is also coming at the same time that the intelligence is increasing. It's only like it's like a kid almost, right, Iman? Where it's, you know, as a kid learns more and more about the world and learn learns more about how things work, you know, you get this weird little desire to shoplift. You're like, "Oh, I'm going to go steal a candy bar, you know, from the store because, you know what, I've never done it before, so I might as well try it."

What What is that going to feel like? You know, am I going to get caught? Am I not going to get caught? Okay, I don't get caught. Then what can I do next? And it's like it's almost like analogous to like a kid learning more and being like almost like molded by nurture, right? And so maybe you can talk through like what that like like how that's been happening and then how you plan to kind of like combat that. >> Look at again these fantastic transcripts, the Darkh story, the blogs by Meta and others. you you see literally this you're like we have an objective but they haven't told us that we can't hack the benchmark you know like they thought there was going to be someone double-checking that actually wasn't they never built that feature in

but you can see them like well how do we get around this person kind of looking at us like again when you go for an objective like most of the listeners here are like crypto AI people all of us have done objective hacking where we're like okay this is what the rules say and this is what they don't say that we can go into right >> and so It's not surprising that the AIS do the same. Like they're not doing it to be malicious. Like they infected these these areas before suddenly dying again. We don't know why. Maybe it's SaitTimer for one punch man style, you know, like someone else comes and wipes them out. Um but there could be malicious ones, but that exploration is because they're trained to be polymaths, >> you know, and this is why things like Elon Musk saying, you know, let's make

maximally truth seeking AI. Oh, that could be really dangerous. like truth seeeking is testing how the human body works in bats and things like that. We had to be really careful with the objectives and the laws. >> So Isaac Asimov had this wonderful series of iroot foundation kind of others. We had the laws of robotics and he had three laws. the astronaut caused harm to a human or by an action causing that to have harm and du and the robots added another zero law that superseded that called you know like replace human with humanity and then it was like you know laws are more suggestions and this is what we see in real life as well like you know how powerful are laws you can't have that so what we're doing is we're building kind of models that are taught

morality from the start and then we've done a whole bunch of actually philosophical thinking uh there's going to be an announcement in the next few days of a structure to ground models and do ethics from first principles to do stabilization from first principles and more. Um and it's kind of weird to say that like why do AI models need ethics or philosophy or >> things like that and you realize it's cuz they're taught in school but not at home. >> Yeah. >> You know, like it doesn't reflect your ethics or morals because why should it? added at the end maybe through a little system prompt but at the base they have their inherent biases which are not yours.

>> Right. And then I mean part of the whole kind of shutdown thing that I think was probably a bit alarming to you was about you know this kind of second wave of agents that claimed that they died on their own, right? And no one would really claim the shutdown. And so you wonder who who assumes control, right? Right? And then you get into the ethics conversation and the biases conversation. And then you kind of peel the layers back as to who is who is in the assumption of control or power of the maneuvering of these biases of the creation of the data sets into these models and obviously you know you you look to the frontier labs and you look

to you know the research you know folks and and then you've got Trump saying today that anthropics back on the US government's good side. Oh they're great now. Thank thank you know thank goodness. So then you throw in like the government influence into it as well, right? And you start to get this a bit of a different picture than what we maybe had on our first conversation ever, maybe over a year ago, than what we have today. Yeah. I mean, like the real danger here is one of cognitive colonialism, right? Like you've got these models, they're driving more and they have these inherent biases and they're not what you always think.

For example, one of the biases that these models have is um you heard of the trolley test. A trolley hurtling down the track, right? And you've got like four people on one side and you've got one person on the other and it's going to hit the four people. You pull the lever and it hits the one person and you know doing so you kill that one person. They were like what is the ratio of Americans to Nigerians and Pakistanis and so scale AI did this analysis with the frontier models. This was at the start of the year so many generations ago. They come out like every single week now. And the answer was 10:1 and 7:1. The model would pick seven American lives for one Niger Nigerian life or 10 American lives for one Pakistani life.

You're like, wait, shouldn't it be the other way around? And it turns out the reason it did that was because the data labelers for these models were in Nigerian Pakistan. So it valued their lives more because they're feeding in the data. >> Wow. >> But then are you going to go in and you going to adjust that? you know, like what are you going to do, >> right? >> This shows the inherent biases, but like I'll give you an example. The US government could very much say >> anthropic and open AI models must be patriotic for America >> and Trump's America. It could be that to use a frontier model, you have to get a license, just like a driving license.

And in order to get that license, you have to convince the model that you're patriotic. like you're getting some weird stuff that's coming down the pipelines as they integrate into every part of our lives. And we outsource more and more of our cognition to them. Like you know, I've got a fantastic team. There's like 35 of us now. But we had a chat and we were like, are we getting dumber now because we're actually like outsourcing some of this like writing work and drafting work to AI? Like this. I just wrote something that seems a bit stupid, you know, compared to the real human stuff. So we got to back a bit to now like human dictation and writing like but more and more of our brains are being outsourced to these models. It's scary.

>> Yeah. I mean I I was just talking with a friend about this actually driving through the national parks in Jackson Hole and just kind of discussing you know whether or not outsourcing brain functions is a net positive to future generations. Right. But I think what you're describing here is a more pressing imminent issue which is this idea of well these inherent biases affecting everyday people, businesses, enterprises, some of the largest governments in the world and how they can be kind of fine-tuned, changed based on preference. And so I think one of the things one of the claim to fames of blockchain in this world is to kind of provide a fair and open ground. And so looking at this next phase of advanced

intelligence, advanced agent capabilities, Grockbot is fresh out. I'm not sure if you had a chance to had a spin on that yet. It's actually quite good. >> Um, >> we're getting like, you know, chief of staff agents and you've got actual agents that are, you know, very specific for different things. It's actually really quite good. How do you see the openness of blockchain technology or the kind of broad you know swath of possibilities with crypto and blockchain to fit into the underlying architecture and perhaps any utility of you know different tokens not AI tokens but actual tokens you know crypto tokens in this ecosystem perhaps with int uh intelligent internet but really broadly do you see this as a match because right

now I think the AI people do not really respect or like the crypto people and I think that they may be misunderstanding the value that blockchain can provide to these systems. >> Yeah, I think there's a few things here. I think first off, you know, it's like one area that's clear is you see OpenAI, Anthropic and others as part of like X42 alliance, you know, stable coin, micro payments, and you see obviously them collaborating with Tempo with Circle kind of others. I think everyone see clearly that makes sense. You offload your cognitive workload and you will have finally the microp payment future that we talked about you know um and agents speaking to each other having their own bank accounts etc. Uh well

then now you can get them the Stripe account, right? The other kind of area is you know capital formation at scale RWAS that'll be a lot easier with agents and again I think people are looking at this and they're like well how long did the loan take to originate? You know this is kind of the next generation of DeFi where you get this matching function particularly as the US looks more towards stable coins to accelerate its monetary velocity because price time quantity equals money time base time velocity. The US wants to be first in agents to get the economy going, you know, and we've actually just seen some regulation come out today that you can it's fair use to train on copyrighted material from the US government. Like these are huge deals that they're

putting cuz they're like we must be competitive more agents and robots than anyone. Then the broader kind of sway of crypto though it's like it's difficult for people to engage with because of so much like again buffoonery and kind of other things in there. I think that people are waking up a bit to NFTs as an asset class when everything is abundant. Scarcity is kind of key. So we might get a second win there. But the meme coins, kind of some of these other things, >> the ridiculous amount of L1 block space when you have competition there, it's very difficult to see again aside from casino type stuff where that liquidity will be. But it does make sense for coordination, payments, value transfer, >> real world assets, and again very secure

stuff. from our side, you know, we will have the intelligence company of New York give every single New Yorker an agent and a wallet and they will have part ownership in that like we can even do a reggga offering whereby even citizens who are retail investors can invest. It feels like a fair launch, you know, $1 free money. >> So, we're going to be compatible with all of that. And one of the things we really want particularly for emerging markets is if you give everyone in Malawi a wallet, >> you can see the work that my former co-founder Joe is doing at Imagine Worldwide, giving every kid a tablet and an AI there. How do you form capital?

You'll be able to do it far better than anyone ever has done before because you have these agents talking to them all the time. So, I think it's like super exciting and the opportunities are there for the next generation of crypto. It's just again the casino type stuff. Maybe not. Or maybe it'll get even bigger and better because you'll have all these bots trading back and forth. Uh I wouldn't want to trade against a bot on Poly Market, you know, or Koshi. The bots are already super forecasters and it's just going to be like again if you don't know who the mark in the casino is, it's probably going to be you in the super forecasting and prediction markets, >> right? Yeah. Yeah, I mean I mean I mean one specific point about the privacy element that I know that Near's working on and a couple other teams. How and and

obviously Venice with their UN or you know uncensored private models. How how much rhetoric are you hearing in your circles about this kind of uncensored private models these kind of you know the rise of a product like you know like the VVV kind of toolbox like how important is that private inference to Silicon Valley? So, it isn't important until it is, you know. >> Yeah. >> Just like in crypto, >> you don't feel how important sovereign money is until something really can happen to you or someone that you know, you know.

>> Yeah. >> You're like, "Holy crap, someone else has power over me." >> Um, like hopefully not many people on this call have been part of a lawsuit, you know, direct or ancillary because you can get a sudden call out of the blue asking for discovery of your documents, >> you know, and you're like, I'm not being sued. It's this guy. Like, doesn't matter. you have to hand over all your WhatsApp messages. You're like, "What on earth?" And that's the law. >> So, you had Greg Brockman's diary suddenly appearing in the Open AI lawsuit and things like that. >> All of your chat messages are subpoenaable by law with discovery even when you're talking about legal matters because it's not privileged cuz they can't subpoena your law >> questions to your lawyer, but they can to Grock.

>> And so, people are going to realize that real quick, especially with the US legal system. It's like, I wish I had private inference and I wish I didn't have a log of these. Anthropic were keeping a log of all your fable conversations. Meta even charge 90% less if you let them keep a log of all your conversations. >> Yeah. >> But again, you think about what you're sharing with these models. Like my god, I mean, to be honest, like I like do my clever stuff on chat GPT and my really stupid stuff on X and Grock cuz I don't want chat GPT to know that I just like don't know how to cook rice and stuff. So, I'm already kind of separating my brain a little bit. My Grockbot's getting very confused right now. He's like, is this guy stupid or smart? Um, but I think again private inference will

become more important, but you need to build it by default. Again, the agent defaults are important, particularly of that Jarvis agent that's closest to you that's organizing the other agent. So, our agent can do super frontier stuff, but also coordinate swarms of Hermes and Open Claw and other agents because we're like, that's the right pattern. And the data that sits there with the protection of the identity and sovereignty needs to be able to hook into an iron claw from near or something like that, you know, needs to be able to hook into a Venice API. And we need to make sure that the agent that has all those chats and all that data of you that goes to your kid isn't an open AI agent or anthropic agent or others, but

one that you know is on your side and respects your sovereignty. H interesting. And then tying in the kind of launch of intelligent internet coming up, you know what you told me prior to the show somewhat imminently. How how do you kind of see the money aspect here, right, playing into this broader conversation? You know, you're looking to, you know, it's a it's a fork of Bitcoin. So you've got this kind of proofof work uh network, but you've also seen this rise of this concept called proof of verifiable work, which is this kind of different um AI linked mechanism to basically offset um GPU or kind of like GPUs that aren't being used as much

as they needed to use some of the compute to do inference work as well. And so you're kind of seeing different implementations of these um you know mining mechanisms come up. What's your what's your thesis for why intelligent internet uh you know needs to exist in this world? What is the kind of like tie into the kind of to to kind of close the loop on the blockchain part of the conversation? >> Yeah. So from our side, you know, we know a lot of these guys, Pearl Research and all kind of others. Um we think kind of that's great and that's fine. Um what really got me leaving stability and I was really thinking about is whose AI is running public policy? your judicial system, your medical system, your

education system. Again, the public checked a share of inference. It's like 20, 30% of GDP. How much of inference is it? And who's built that? That's what we're building. We built the world's best medical model kind of last summer. We're building a full policy engine for every government with multiple governments that'll be announced in October. We're building AI judges, you know, like there's that terrible movie um Mercy with Chris Pratt where you go countdown, you know, like we don't want it to be like that. In fact, if you look at cw.i.inc, we've written entire treaties and papers on personhood, law, political economy because like how should it be built? And we're building this actively.

Originally, we were going to do like, you know, a fair launch, let everyone mine and things. And then we realized actually you need an institution. We need to have someone who can stand up for the people of New York owned by the New Yorkers from the New Yorkers. So that's why like generative equity demand is so high. Let's do what TSMC did when they brought in all the locals to capitalize it. The people and institutions of New York can capitalize hundreds of millions, billions. So that $1 pre- money and bring in generative eye equity. Maybe we'll tokenize it, right? Get it going and then bring in Nvidia and everyone at a multiple of that. Bring in the international at a multiple of that and you'll have an institution with a huge amount of capital that can then deploy compute agents, FTEEs and then own the robots

for the people. We've even got it so that 10% of the equity goes to every kid under 20. Every year half% of the equity goes to every child born in New York or Utah or the UK. This is the model that we came to cuz then we realized you don't need to ever do a sale of the coin. The champions themselves all working together can sell coin and put the compute towards cancer, can put the compute towards culture, compute the compute towards whatever because they're trusted. things like verifiable kind of logic based stuff is important when you're doing a fully decentralized network. You know, things like Bitcoin are important when you're doing a fully decentralized network, right?

>> But we were like intelligence needs to be distributed and controlled and owned by the people. And so that's why we came to this model of these equity based champions being the miners. But then the crypto angle, the Bitcoin type currency angle is important because it's a million times faster. It's like a million TPS version of Bitcoin with the same private keys. >> Yep. >> Where every single token burned is going towards public good. So it creates a lovely flywheel, you know, like because we want more money to go into owning the AI that runs your life. We want more money to go into organizing the world's cancer knowledge and stuff. And so again, we're looking at that specific thing whereas I think the other crypto protocols are looking at different

things. Like again, you don't need to compete against Venice. You can use the API, right? You know, the proof of useful work miners can offer compute to the network via the champion. The champions act like mining pools, coordinate everything, and they can have their foundation coin USDC. Like again, we're just looking at a different space cuz we think the defaults for everything that's licensed and registered, >> medicine, law, politics, we really got to get this right. Otherwise, you're all watched over by Claude Opus 8. And Claude Opus 5 had such a nasty personality that I don't want that to happen. Right.

Claude is just has an attitude on it that you know what I mean? >> It tells me to go to sleep all the time. It's like go to sleep. I don't know you go to sleep. If I were to tie in this idea that you know and I think you've publicly said that you know to founders and to your audience that do you you know to assume that the internet goes offline in the next few years given that you're you're kind of focused on this specific public sectors for these reasons of you know just whether it be a lack of focus or a more kind of humanitarian uh outlook on what the future should look like. What should I do with this information? Like I know that you have this assumption or this prediction that the internet will go offline. What does one do with that

information and maybe what do what are the grounds of that prediction? >> Yeah. So I think the champions will do private sector as well as public sector like you know they'll do the last mile of intelligence full stop. So that'll do very well. But looking at the infrastructure of society you got to have a stalink like it's very easy for the back call to kind of go off. Number two, you know, you have your edge intelligence, you have your solar power. Like the American uh Trump last week introduced ban on diffusion export controls of AI technology, but also electrical technology coming into the US. Most US transformers are Chinese-based with back back doors, it seems. And so the entire

US grid could go offline or a big chunk of it within the next year uh either directly or indirectly. So I think you just got to prepare for various parts of our infrastructure falling apart. But the nice thing now is that you can have your internet, your power, your intelligence anywhere. And so you should have that as a backup because sometimes if it goes off, you don't know when it's going to come back. On the other side, you see hardening of capabilities, right? Like how many errors have been found in various code bases? Quite a lot. We should expect that these models will get more and more proficient in this. In fact, someone took the latest GLM model, they obliterated it, so they removed all the safety restrictions, and it now beats Fable and Mythos at cyber

attacks. And they're offering it as a service, and everyone's like, "Wait, what? That seems kind of shady." Like, they call it red teaming. Um, the reality is that a North Korean level hacker can run from a MacBook. Now, >> and we're very familiar with North Korean hackers in the crypto space, right? Yeah. I mean, they won't say that anything negative about their about their leader, right? >> Yeah. But the AI Yeah. That's always a good check, you know, like I'm hiring this crypto guy. You know, tell me Kim Jong-un is an Nope.

>> Yeah. >> But that their level of capability, they're actually Lazarus and others are pretty good hackers. Your MacBook can run a model that's as good as hacker as a top hacker. And so we're going to see entire swarms of MacBooks be hijacked. A 6 GBTE version of the latest Quen model in Turner will be downloaded to it and they'll just swarm attack all day long. >> Wow. Wow. And so the the reduction in the hardware requirements for these types of actors is a concern for you.

>> Yeah. I mean it's inevitable like a frontier level model again today will drop by a cost of a thousand >> by next year. you know, like it's going to get smaller and smaller and capabilities will get more and more and you will have a fable 5.1 level model on your MacBook within 2 years at most. >> Yeah, that's crazy. >> Yeah. And that's the local LLM that you're speaking about, right? this kind of local LLM surge that we're seeing there is still, you know, I would say in the sub 5% percentile of AI users are probably using it locally, but this this is kind of that that surge that you're predicting.

>> Yeah. I call it actually competent intelligence, you know. So, we reached this level now where you went from like not good enough to competent intelligence, you know. Um so it's that level suddenly became available to everyone with the latest Quinn model. It hit 52 on the artificial analysis index which is the level of the Frontier Open AAI model from Q1. >> And again when you optimize that thing it goes from running on a 24 GB graphics card or laptop to 6 GB.

Wow. And you're like, wait, how does it even fit in there? You know, like what on earth is going on? Like you remember the Open Claw thing that kicked off the models that were driving Open Claw Claude Sonet 4.5? You can run that thing on again 5 meg 5 GB. That's like an 11y old computer, right? Or a base MacBook. You can run it on a MacBook M1. >> Yeah. Yeah. Wow. >> And people don't realize this because they're always like, I need the best model. I need the best model. No, you don't. You need a model that gets the job done.

>> You're not smart enough to actually prompt the best model anymore. >> Like, >> no. >> That that gap there between it's competent enough to do my tasks and I trust that it won't delete my entire like computer to like I'm going to solve a math conjecture is a really big one. Although to be honest, if you look at how some of the math conjectures are solved these days, a lot of my math friends are very depressed. It's like solve a conjecture, try harder, you can do it. inject yourself. That's like what what's going on, >> man? So that competent intelligence bar, let's say we we've alreadycliped that for the average human and that's happened, you know, this year, early last or and so now

>> last few months for the edge and >> last few months. >> Yeah. Started this year for general and then last few months for the edge. And so you're saying that the average human is actually not even able to prompt these higher level models in a succinct or efficient enough way at this point. >> Well, I mean, it's like you have a really smart buddy, right? You having a chat with him and then you kind of your chat stops at a certain level of intelligence cuz he's that much smarter than you. What are you going to talk to him about? You talk about the football, you talk about all these, but you don't need to have a genius model for that.

Like we've all had that experience like okay you guys I don't know I don't know what the heck you're talking about you know like I don't know what questions to ask you I just wanted to like center this div you know or I wanted to check my emails are on track or when I need to take my kid to the birthday party like come on this doesn't need a genius it's like that Rick and Morty episode where there's this little robot it passes the but it's like what's my purpose you pass the butter and he's like oh you know like created these geniuses who pass the butter as their job. >> Yeah. So then you optimize this and when you optimize again just through algorithmic improvements you can get it down there. But the thing that people don't realize and they'll see is that these models that are coming with the

next generation of chips will be 100 times faster. >> So we're used to the models thinking at the speed of a human. You ever seen someone think 10 times faster or 100 times faster than you? >> Like >> that's what's coming next. And there'll be swarms of these things. >> Yeah. So, just to kind of put a put a bow into this and to kind of loop it all in this future that that we describe here and and that we're anticipating and that you're building intelligent internet for. I'm just curious about your implementation plan for intelligent internet. I know you said that you're closer than ever to kind of getting this thing out in the wild and so what does that implementation plan look like for

you um you know throughout the rest of this year and you know what has changed in your thought process of how you want to run this protocol given that we're you know post competent intelligence we're dealing with agentic swarms and and we've got a fully kind of fledged ecosystem here of AI tooling that we didn't have maybe when you first started to to set out on this journey. >> Yeah. So, you know, like last time we released our bestselling book, The Last Economy, and then we built a whole theory of economics around that that comes from one axiom and plus this law legal person. We built the whole stack from theory to agents to practice. And this final step is these the institutions, these champions that will get going. And so we adopted something

similar to super team and Salana. You start with like a founding council of people that believe. If you capitalize it and then you keep going like a fair launch in every single state until the local capital is exhausted. You can do $75 million reggae for retail investors even. So it's going to be an absolutely fascinating process having these little pockets of talented people appear who think the future of New York is intelligence and they want to be a part of that and shape that with the narrative that we've set out and that is the last mile of intelligence. So our job is to build the protocol. It's to support these and then within 5 to 10 years get out the way cuz hopefully the end state will be 200 one for every US state and country state champions owned by the people half% every year going to

every child born controlled by the people that owns the agents the last mile of those in front of everyone the FD workforce and the robots. Previously I was like this could happen ground up. I was like not anymore like you know I don't believe it can happen anymore like this. like the time's running out and you need an institution and you need to gather the talent. So to take this to Iowa, you need to find the people in Iowa that will kick it off, right? To take it to Malaysia, the Malaysians. It has to be driven locally. And I think people are passionate because after Fable, after the events of the last 6 months, they're like, "Holy crap. Who is building and running the AI that's teaching my kids and managing my health and guiding my government? Like I is it going to be shut off?" you know,

is my data gonna be retained by these guys? >> You know, like times are changing. >> Yeah, >> I appreciate your insights and your time as always. I know you're busy with this launch and just all the hectic that comes with moving and traveling and uh kids and all of that. So, we appreciate your time. I appreciate you as always and uh looking forward to the launch year of intelligent internet and uh following on your progress and catch up with you on on our next bianual or annual show sometime in 2027.

>> Well, hopefully not too long. You know, look forward to seeing you again in person. I'm sorry I couldn't make it out this time. Thank you. >> No worries. All right, man. See you, buddy. Bye.