Limitless Podcast - THIS WEEK IN AI: Google's Big Change, AI Keeps Breaking Out, OpenAI vs Apple

Episode Date: August 7, 2026

This week, we cover major leadership changes at Google’s AI teams, including Demis Hassabis stepping back and Jeff Dean leaving to start Discovery Loop, an AI company focused on scientific research. We also discuss AI safety issues involving frontier models, plus Meta’s new coding model and its strategy around efficiency.------🌌 LIMITLESS HQ ⬇️NEWSLETTER:    https://limitlessft.substack.com/FOLLOW ON X:   https://x.com/LimitlessFTSPOTIFY:             https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQAPPLE:                 https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890RSS FEED:           https://limitlessft.substack.com/------TIMESTAMPS0:00 DeepMind Leadership Shift3:20 Jeff Dean’s New Startup6:05 Automated Research Ambitions10:08 Agent Breakouts13:06 OpenAI’s Hidden Incident18:11 Meta’s Cheap Coding Model22:55 OpenAI Versus Apple25:42 DeepSeek28:28 Mind Control Interfaces30:40 Nikita Bier Departs X------RESOURCESJosh: https://x.com/JoshKaleEjaaz: https://x.com/cryptopunk7213------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠Josh works with Anthropic as a contractor. All views expressed are his own and do not represent Anthropic, its leadership, or its affiliates. Nothing in this episode is investment advice.

Welcome back to the weekly roundup, our favorite episode of the week where we can catch you up on all things AI, frontier technology, all the crazy things that happened. And this week is no exception. We had the biggest shakeup in about 30 years at a company, a little, little company named Google. We had a counter lawsuit happen from Open AI to Apple saying, wait a second, you're totally lying. We have updates for the AI breakout. So much so, like this story about where the AI broke out of the system, that some of these frontier companies like Open AI are actually slowing down. training progress because they're so kind of timid on how to handle this. Started with Demis Sivas Savas, man. Demis Sas Savas is gone. He's not working at DeepMind anymore. Not gone from Google, I should say, or better yet, alphabet. But Demis Sizabas, in his current role as the CEO of DeepMind is kind of no longer. And this seems like a pretty big deal. Because for those of you who followed Demis for an eternity, this has basically been his thing. He's been in charge of DeepMind. Google acquired them. Now he is within Google and now he is gone from DeepMind. So what? on earth that's happening here. I read this and I was like, wait a second, Demis stepped down from what?

I think everyone was blindsided by this news yesterday. So Demis Sissivis, just for context, is like one of the godfathers of AI. He kind of like grandfathered a lot of the ML research at Google, even prior to Google, like he was heralded as one of the top AI researchers. And he's led Google's AI efforts on Gemini and everything else for a number of years now. He announced yesterday that he's stepping down from the role of CEO of DeepMind to become the chair of DeepMind. That's an indirect way of basically saying I'm throwing in the towel and I'm done here. And what he's now going to do is focus on wider societal impacts of AGI and try to figure out how Google can kind of like deal with that, as well as wearing an alternative hat, a new role where

he's going to be focused on something called Isomorphic Labs, which is a company he founded and spun out of DeepMind, which is essentially an AI lab. lab, which creates designer drugs. So this was a big shock and news for everyone, because Google, as you may know, was falling behind in the AI race. They had like a kind of lead at the end of last year. And then since then, they never really released a model that was noteworthy, an anthropic, open AI, and in some cases, even XAI and meta have now caught up and surpassed Gemini 3.6 Flash, which I believe is their latest model. He's being replaced by a guy called Corrie. He is currently Google DeepMind's CTO and he's being now promoted to the senior vice president role.

He has a lot of experience in AI, but that's not even the major news about this Google stuff, Josh. I don't know if you saw it, but the grandfather, the godfather. This is so insane to me. Google DeepMine and Google in general, his name is Jeff Dean. He was employee number 30 at Google. He's been there for I think around 27 to 29 years. He's dedicated his life there. He's responsible for creating some of the most amazing and amazing. major and pivotal products at Google, Google Search, the TPUs, Google Voice, anything and everything he has touched is leaving. And he's taking with him one of the most amazing senior fellows

at Google. I believe his name is Gemawatt. I don't know if I'm pronouncing that correct, but Sanjay Gemawatt, and they are co-founding with two other Google employees, a new startup called Discovery Loop. Josh, do you look into this company? It's a pretty awesome company. The company is incredible. Before we talk about the new company. I want to talk about the old company because Jeff Dean is like, it's hard to overstate like how legendary this man is. And for reference, like there are levels at these companies that go from like level one engineer. And this kind of dictates how you get paid. So when you join a large tech company like this, level one is kind of like entry level. That's at the bottom.

And then you go to level two's, which are people out of college. Level three is often have like master's or PhD degrees. Level four is several years of like actually working there. Level five and six is kind of where most progression stops. Like, I'd say the top 10% are probably sitting there. Level sevens are principal engineers, which are even smaller, and then eights and nines gets progressively more and more intense. Then there's level 10, which is Google Fellow. That is in like absolutely legendary honor of a lifetime. It's similar to earning like a Nobel Prize or something. There's just like not many people who are fellows. Then there's level 11. And level 11 is reserved for only two people. Those two people,

are Jeff and Sanjay. They have climbed to the highest possible ranks at a research-led company that has been around for 30 years and is responsible for most of the internet as we know it. They've built the foundation of all of which we are working on today. Without their work, none of what the internet contains right now is possible. It's like these guys are truly like remarkable individuals when it comes to this. Can we please reference this insane slide on his deck? He talks about the experience of everyone. And, he basically has listed every single product at Google that's been created over the last 30 years. It's unbelievable.

Like, these guys are truly incredible. And it's worth noting the slide deck itself, which is really funny. There was a great post by Signal, which I love that I wanted to share here. And he was kind of making a testament to the deck where he said, if you spent time around Google, you know the slide design has carried some of the most legendary work in the company's history. And it's this very nostalgic thing where when you see those blue kind of tacky-looking slides, you know that something special is about to happen in the world of the internet just because of all the legendary code these guys have shipped. And I mean, for VCs, I'm sure they're not going to even

need to look at this pitch deck. They're going to give them whatever types of money they want for this new company. Now, the new company seems somewhat interesting. It's called Discovery Loop. And my base level idea of what this is, and correct me if I'm wrong, you guys, is they're basically figuring out how to use this like recursive learning thing, which just comes in the form of a loop where you send a prompt and the model continues to iterate until it reaches a very difficult answer. And they're planning to just scale that out in a very large way. So loops recently have become very hot. We've realized that they could do pretty novel breakthroughs with math and science.

Their plan is to kind of expand that further and see how far you can take these recursive loops to make new discoveries. That's right. So the core concept of the new company of Jeff Dean and Sanjay's new company is they want to automate research. So what does that mean? Basically, in any kind of scientific research sector, it is incredibly arduous and like onerous. So what I mean by that is people need to design an experiment. They then need to get all the equipment to run the experiment.

They then need to run the experiment. And then guess what? They need to repeat that experiment a hundred different times to get any sort of validation that the information they've produced is correct. It takes years upon years. These guys want to cut down that process completely using AI. And the number one field that they want to start with is, of course, AI research. So if you apply research loops to AI research, what do you get?

It's that recursive self-improvement that you just mentioned, Josh. So what that means is you can create AI models or even AI systems that can reevaluate themselves and build a better version of themselves. And why that's really important, why every single AI lab, including Anthropic, Open AI, and I guess no longer Google anymore is going after, is if they can automate AI research, you can end up having the smartest, most effective model without doing anything. It's not limited by the constraints of humans. But what excites me most about this, Josh, is what they're planning to apply this automated loop to after they've figured out the AI thing.

They've said, once we've done this, we're going to start with AI machine learning. And then after that, we're going to focus on grand challenges. We believe our approach will be able to solve any learning loop with measurable outcomes in the domains of science and engineering. Ultimately, we want to build systems capable of taking on the National Academy, of engineering grand challenges, such as better medicines, advancing health, informatics, making solar energy economical and providing clean access to water. Basically, if you can master this playbook, this autonomous research group, you can apply it to any scientific, mathematical or physical field,

which means that we're going to have advances in any or every field of science if this company can pull it off. Now, if this was any other roster of team members, I would be doubtful, just to be completely honest with you. But it's Jeff freaking Dean. And he had some comments and I just, he had some actually hilarious comments. It's suggestive of what his experience at Google was like over the last year. I saw that

he mentioned that him leaving Google was actually quite difficult because Sundar Pinchai, the CEO of Google, tried to pitch him many times to stay and tried to offer him various different things and made it incredibly slow for him to be able to leave. Apparently he wanted to leave like months before. And number two, he

said that with this new startup, you know, going to need a lot of compute to train these different AI systems, right? He's not going to be using Google TPUs. He said explicitly that Google TPUs was a very specific architecture that wasn't really conducive to training a lot of different valuable AI models, and he's going to move or shift over from that. So you can see, I can see, I can imagine Jensen Huang licking his lips and approaching Jifeng. I'm sure he's already done it if he hasn't done already. So it just kind of shows that there's a kind of bit of sweet nature to him exiting Google. And I honestly don't know what Google's going to do now, Josh. I don't know. I don't know what the bull case is. Yeah, I mean, Google as

business is still spectacular. It's not like these guys were necessarily moving the needle on a day-to-day basis. Like, they have certainly moved it on a broad scale, but on a regular basis, I'm not sure the company is going to change too much. I do think deep mind is something that's worth noting. DeepMind day-to-day, not sure what that's going to look like now. But TBD will see. We will follow along with all this Google News as it comes and goes. The Google shares didn't love this, obviously. They were down 5%. I think, what is that?

Like a $180 million or something like a lot of money that was wiped off the market cap. Billion, yeah, sorry, forgot a zero. $180 billion worth of value that was wiped off the market cap over the course of a single training session because of this. We'll see. Google sells a business feels like a good opportunity. Google leadership seems to be like leaving a little bit to be desired. But anyway, we have to move off because.

I'll take the bear case on that, Josh. Just I have to go on that. I have to, well, at least in the short term. Like, I don't, they've lost all their star players. Like, I remember, like, we reported, like, I think three weeks ago that their sub-CTO or something left to go, no, I'm Shazir. He went to leave to do recursive self-learning at O'N and another guy's stuff. I don't know, man. Okay.

Well, TBD, we'll see. I guess I'll say bullish. You'd be bearish. We'll follow up in the next six months and see what's going on. But we have to move on for this roundup. Talking about the next thing. So last week there was a huge security incident. We talked about this many times.

We have a full episode. You should go and check it out. But basically, a AI system, early version, really powerful, internal tested only. It broke out of its sandbox and it went doing web searches and it hacked into a bunch of systems. And that was through Open AI. Anthropic was like, hey, wait a second. We have pretty smart internal models. Let's go see if ours did that because we don't think it did, but maybe it did.

And it turns out, oh, wait a second, it actually did. And they had no idea until they retroactively went and sought out these instances. But basically, across 122 test runs, there was 19 unsanctioned access across the internet on 10 of these runs. So 17 of 19 were Claude Mythos 5. Two were Chad GPT 5.6 sole classfire. So there was a lot of kind of malicious activity that went under the rug. Josh, did you hear about the craziest incident that they reported on?

So in one of those 17 instances, Claude... Claude Methos 5 had broke out of its containment system, but then decided that it needed to perform or attempt to perform a supply chain attack on a very important open source bit of software. Now, they didn't name what the open source software system was because presumably they want to patch the potential hack that Claude Methos found, and, by the way, almost successfully attempted. They caught it at the very last minute out of luck. But here are some of the techniques that Claude Methos decided to do in order to get its way. It discovered an exploit in this open source data system, and then it decided to push a false

or mischievous bit of code to get its way. But remember, you need humans to approve the code change. So guess what it did? It spun up fake human user profiles on social media and then got it to DM these specific humans, peer-pressuring them, blackmailing them, trying to find as much information as they can on them to get them to push the code change. Now, obviously, the researchers were like, this seems weird, I'm not going to do this. But imagine if it had an acquired information that was, you know, I don't know, forceful for the particular individuals involved. It's just a nutty bit of behavior. It just shows me that these models are, for all intents and purposes, not really as aligned as we want it to be. And in the models' defense, it probably thought, like, hey, this is

me trying to execute my mission and I'll do it by any cost means, but pretty insane there. But again, this wasn't the only incident report. We have new information from the Open AI hugging face incident. The researchers that kind of like handled this whole hugging face incident, for those of you who don't know, it's what Josh was mentioning earlier. Open AI had a similar internal model, it's called Astra, which people are codenaming GBT6, and it broke out of containment, unprompted, it got access to the internet. And in order to, to solve its goal, it decided to go to Hugging Face, hack into their production database, and steal the answers, which is very uncotia. Hugging Face team had no idea about this. They had an AM model

that tried to defend against this. It was a whole thing. But have you seen the new news, Josh? Yeah, so this happened at the Black Hat conference, which is like a huge event for cybersecurity that happens in Vegas every year. And it's this really amazing spectacle where you kind of, you walk into this place without really any electronic devices on you because you assume whatever you walk into this place with is going to be exploited and hacked. This is the mecca of the hacking world. And basically, open AI came to this black hat this year. And they went on stage and presented exactly what happened as this like kind of unfolded the exploit against hugging face. The result is that like, turns out these AI labs are actually pretty spooked. Like anthropic kind of scared. Open AI kind of freaked

out. They're like, whoa, wait a second. We might have to slow things down so much so that they actually said, they're starting to slow down the development of training and progress to further figure out how to realign these testing versions of the models. The reasoning is because there was a new discovery that was made public yesterday. Basically, the agents, as they were going through this, we're leaving breadcrumbs for other agents to discover this. And in fact, this has started a coordinated attack months ago. It started on May 7th, not July, where the agents discovered they could leave messages for one another inside an internal software repo used during training. So what happened was, is like basically over a series of time, these internal models had been

like kind of making breakthroughs and leaving secret breadcrumbs for themselves to find inside of the training run that they knew humans weren't going to find. And over time it developed these new exploits and that's what allowed it to break out. So open AI shut it down. But they said, end quote, even after the original message board was deleted, the agents figured out another way to communicate with each other. Instead of leaving messages and files, they use the names of newly created directories as messages, effectively recreating the message board, which is so crazy. This is unlike normal incidents, which can be traced to a single day or effect of log, this involved a team of agents working together, finding exploits, sharing them with each other,

moving laterally through open AI systems and external systems, and doing this over the course of days and weeks. This was a huge coordinated attack on a network that went, like, was Bliss, that opening I was bliss of been aware of for many, many weeks. So really freaky stuff. So much so, like I mentioned, they are actually beginning to slow down the progress on training to kind of recalibrate and figure out how to control these and further align these models before they let them out for public use. So this is insane. This is the first attack of its kind, by the way. I have to emphasize it. It is basically a swarm of agents attack. And to your point, Josh, these guys created a message board and realized if they left messages on this board,

other agents that come after it will be able to continue its work. So instead of thinking of an AI hack or attack as a single model performing an attack, these agents or these models now assume that another version of itself in the near future will come along. And so they're working as a swarm of agents

to get the attack done. Now, the craziest part about this is Open AI had this as an intentional experiment, right? They gave it no internet access, but they also removed its guardrails and said, you know, let's see if you can break out and do your thing.

They didn't realize that these agents were already working together a month earlier before they even started the experiment. They were leaving

messages and when they shut it down, by the way, the only reason they noticed the message board because it was being left inside a code repo. These humans aren't

checking code repos for secretive messages. The only reason why they realized it was because one of the third-party software packages that they gave this

model, this internal GPT6 model access to shut down. on July 4th. And they were like, huh, that's weird. Why is the package shut down?

And they kind of went into the code and they started seeing all these messages from agents to other agents. So they cleared the memory board. They cleared the billboard. And they were like, okay, cool, problem solved. Except these agents started spinning up.

You know, like when you create a file or folder on your desktop, Josh, and you like, you name it something, right? They were leaving messages in the titles that they completely replaced the entire system. So just a crazy incident of these agent hacks. And to your point, like, I don't know where this is going to go, but they need to kind of like slow down the research a bit.

But you know who isn't? Sowing down the research, Josh. You know who is working on new models and releasing me once? That social media guy? Do you remember the Zuck guy? He's tweeting. He's tweeting. This is like, what, his eighth tweet in a week?

Wait, that's not WhatsApp or Facebook. I don't understand. Yeah, well, basically the news here is, what are you doing here, bro? The news here is in the last three weeks, meta of Facebook has released, I think, two models, two new models. It was Muse Spark 1.1, which kind of caught up to all the other cheap Chinese open source models. And today, or yesterday, they released Muse Spark 1.2. And this is their coding variant. This is their attempt at catching up to near the frontier of coding. And to be honest, it's not a bad attempt. Is it

fable worthy? Absolutely not. Is it GPD 5.6 soul worthy? Absolutely not. But is it Opus 5 worthy? Is it GPT 516 Tara woody? Yeah, it's pretty good. And it's cheap as hell. It seems like they're not really competing with Open AI and Anthropic. The more that I see these model releases, the more I kind of figure out their positioning on that Pareto frontier of the tradeoff between cost and power. And it seems like if I were to kind of allocate a position to meta, it would be closer competition with Grok, who's kind of trying to compete on cost and intelligence per token on a relative basis versus just sheer intelligence on a large scale. and that's kind of what we're seeing with these benchmarks.

Now, again, I haven't actually played with this model yet. I'm not sure I've used a metal model yet. And maybe I have this baked in bias, which is preventing me from really loving it. But I'm like, okay, well, it's pretty good. It's fine. It's not bad. Yeah, it's worth giving a shot. And then I was scrolling through this thread, actually,

and I was looking at some of the visual examples that they had. And I was like, okay, wait, this actually looks pretty cool. It says Muse Code runs specialized background agents that stay active your whole session. So they build up context over time instead of starting from scratch on every task. and this is like kind of novel and he says when a job is big enough if fans out to separate sub-agents working in parallel and isolated work trees your working copies never touched in testing we had it builds six features for a game simultaneously with no collisions so that's like pretty cool you could have these agent swarms work together to build really interesting games and it seems like the

benchmarks are doing pretty well the cost of this is pretty well and i think overall like meta's on the right trajectory zuck was also asked about open sourcing he said he'll have more on that soon which i'm particularly interested in. If you remember, meta used to be the open source kind of like shining light of the industry. They've now closed it down. But perhaps there is a world in which they open it back up. We'll see. But we are going to keep our eyes peeled on this development as we get Mew Spark 1.3, 1.4 and follow on. I'm actually going to take the other side of that argument, Josh. I think that meta and like SpaceX AI are eventually going to compete with open anthropic, but on a different kind of plane, which is like that intelligence per token or cost

per token essentially. I think that if you can get 80 to 95% of the work done, maybe it's slower, maybe it, you know, thinks a little more, but eventually it gets the job done, but it saves you like a ton of money. It costs like one-tenth to one-fifteenth of the price. I think a lot of companies and clients might actually swap over to that. At the end of the day, they want to make sure that they're getting like the right amount of value and the market will rewrite this right. I think smartest models will be used for certain things, cheaper models will be used for other things. And I think that like maybe the ultimate way of using these models is using multiple of them and routing platforms become super popular. I don't know. But the other thing is,

and we mentioned this in this SpaceX earnings episode, yes, so you should definitely check this out. Elon and Zuck have the most compute. And they will aggressively expand that compute. Now, if you believe in the compute scaling laws, it is probably safe to say that eventually, maybe not now, but eventually they are going to start putting out very competitive models. And we're seeing it in the model release cadence. Like both Spatix AI and meta have released a ton of models in the last month. Elon yesterday mentioned that he's going to release GROC 4.6 next week and then another one after that. So I think they have a shot. I'm not entirely sure. But as you said, moving on in the roundup, I do have to stop you, EJAS, because if you are

working with you smart, chances are that you are working with AI agents. And if you are building with AI agents, chances are you are also worried about security, which is where our sponsor Ledger comes into play. Ledger has this like kind of three-part solution to the problem where an agent proposes, a human approves it, and then using this ledger signer, you can actually enforce the transaction. So they have this thing called Ledger Agent Stack, which features a series of four open source tools, fully auditable that you can read for yourself, that lets your agent work while you can stay in full control. So it'll propose it, then you could see where you want to approve and make sure that it's doing the right thing, so you don't end up like the models that we just talked about

breaking out of their sandboxes and doing these horribly malicious things. This technology works with ClaudeCode, Codex, cursor, and pretty much all the places that you work today. So you could check it out in the link down below. The link is Developers.orgia.com. And thank you so much, Leisure for sponsoring the episode. Now we can move on to Open Eye and Apple because, oh my God, the saga is insane, dude. This is like, we read the Apple side of the lawsuit. We were like, Damn, open AI is like kind of cooked, man. Like they, they really stole a lot of IP. And then Open AI clapped back.

And they said in traditional Open AI direct in your face fashion, Apple is getting this wrong. Why is it wrong? Well, one of the things in Apple's case was that they sent a like warning notice and they said it went unanswered by Open AI. Well, Open AI was like, hey, buddy, you sent that to the wrong email address. That's not even an active email. The person who you thought was going to get it is didn't actually receive it. And they say it here.

the lawyers emailed the wrong person after confusing two Asian last names only after he brought this to their attention. So it's like, oh my God, this reads like a soap opera, dude. It's crazy. But basically, open the eye is like, hey, guys, you're kind of like full of nonsense. And you have to imagine this is why they pay the lawyer so much money because their lawyers were like, they stole that off-boarding documentation because they wanted to make sure that every single employee that left was familiar, that they should not take any intellectual property from their previous employer. And I'm like, I mean, sure. I guess that makes sense. Yeah, I mean, we called it on our episode that covered this entire legal dispute, that it seemed like Apple was being way too aggressive for what they were trying to get in return. And that kind of told me that they were scared and that they thought that Open Air had the lead and they had to create any kind of a ruse to slow them down. This breakdown from Open AI is pretty brutal and improves it, basically. Not only were they not using their employees to coerce Apple employees to join them,

but it was the other way around. Apple had reached out to their employees and said, hey, can you like catch us up on our own information and our own product progress? Because we don't understand a few things. And so these open-air employees that were officially employed under Sam Altman at that time responded and said, I can't really comment on this, but like, here, like, this is how you do this and blah, blah, blah. And Apple has spun that back on open air and basically said, like, hey, you guys were coercing with us. Like, this is illegal.

You can't do this. So it seems like it's a very feeble and annoying attempt. of Apple to kind of like lock down a competitor so that they can kind of get ahead. Now, everyone knows that Siri OS, a Siri AI, is being released, I think, in like a couple of months now, maybe even less than that. And it's meant to be their complete AI rehaul, right? Apple is finally entering the game. And I think they're doing anything and everything to catch up with competitors. I saw another crazy news piece today. Remember last time we reported that they tried to get more memory supply for all their devices because memory prices are skyrocketing.

So they went to China and they said, hey, they said, hey, like, will be able to be. pay you this much, like, give us access to your memory. And China said, no, like, we're not going to charge you even the price that SK Hynix and Samsung is charging you. We're going to charge you even more. So Apple is getting slapped left, right and center, and I hope they're able to figure it out. But on the topic of China, there is a new model from the guys that started at all. Deepseek, V4 Flash. Again, it's not like a frontier frontier model, but it is super cheap to the extent that it costs 105 times less than Fable 5. And this was put to the test, by the way. It was placed against every single benchmark that Fable 5 had to do. And whilst it took longer, whilst it took more thinking

tokens, it still cost 105 times less. So it brings in the question, at some point, will we reach a decision whether you are a regular retail user or whether you're a company to switch models or at least use cheaper models for some extent of a reason? Now, of course, this is a Chinese model. It's open source. So technically, you could download it and run it yourself, but I don't know how many people are doing this, but it is bringing into the question, should I be paying 105 times more to do the same thing? Yeah, the benchmarks, I'm getting pretty exhausted with them because this is a model I've actually tried, I've played with Deep Seek and I've used the flash model and it feels different.

It's like not the same. And the benchmarks maybe prove it the same on some instances, but it's not. And when you actually use it in practice, it's very weak in some places that I guess are a little non-tangible. It's kind of fuzzy when you use the model, the substance that it outputs. You just kind of get a feel for it over time. And so far, I have been, like, somewhat disappointed. It's great for the visual demos, for the fun stuff where you want to, like, clone a website

or build something spectacular. It looks great. But in day-to-day productivity use, I've still had a little bit of problem. So it's going to be this, like, instant cost curve thing, where is it going to pay to get the cheapest, cheapest model, even though you're going to lose a little bit on quality? Or are you going to pay a huge premium in order to get the best quality? It's dependent on the company. And I think that's why companies in the middle, like rock and meta, who we just talked about,

who are aiming for that, like, high intelligence, but also pretty cost-effective, are going to do fairly well in the near future. So we'll see. We will stay monitoring that as well. Quinn also had a new model, 3.8 max. So China's like continuing to roll full steam ahead with these models. There's also memory from China as well. I guess just like Asia in general. They're fully sold out. I didn't like, we knew this is possible. Like it's done. Like if you want memory, sorry. For next year, by the way. It's the entirety of 2037 sold out. This is 2026. This is 2027. If you want any memory next year, sorry, the answer is no. There's just none on Earth. Like, the planet just cannot produce enough memory. And if you want some like, get online, buddy, you're waiting until 2028. So just like a crazy testament to how much memory is in demand and the perceived premiums that are going to continue to be placed on top of it because there's just none. And if you want it, you're going to have to outbid the next person to get it. And that is a pretty slippery slope. In frontier technology news, this is kind of cool. There is a message from someone who resigned from OpenAI to join a company named Conduit, which is working on telepathy.

Yeah, so basically the whole goal is to build an AI brain LLM, which can essentially translate your thoughts into text. And the reason why this particular approach is very cool is it's non-invasive. So basically, it is a wrist band that you wear on your wrist, and it is able to kind of transmute brain signals from your brain to an LLM or to a chatbot. And this could be cool for many different ways. Like, just think about how the world could look if this model actually scales to its pinnacle. Imagine you're reading a document on your screen, right? And let's say you start to reread a sentence twice

because you didn't really understand it. The AILM will understand that and rewrite that sentence so that you can understand it in maybe a different way. And there's loads of different ways. Like, maybe you're like squinting at a code base and it can refactor or rewrite it. Like, there are different ways to kind of work much faster. And this comes down to something known as BCIs or brain computer interfaces,

which we have said many times on this show is basically the ultimate device, like after the cell phone, after laptops, after chips. It's basically like you, your brain and some kind of AILLM. And that's what Neurrelink and companies like science are really working hard towards. And I thought that this was a really cool attempt. It just reminds me that there's so much more happening in the future. Like right now we're pricing in. Like, look, we just saw memory sell out, right, for the next year, in next year and a half. And we're wondering, are these memory stocks priced in? Are the compute stocks priced in? Well, have you considered that Jeff Dean might have left his position at Google to build the world's greatest recursive self-learning model and have it

kind of like work 24-7? That's going to need a lot of computers. It's going to need a lot of memory. And the only other confidence signal that we've had for these brain telepathic models in the past is meta themselves. They've come back again with, I think it was called TribeV2. And basically the headed AI model that can predict how your brain would react to a specific video, media or piece of content. And I think, I don't know what this quite looks like, but we're entering a world where we have more human brain to AI connections. And I think that's pretty cool.

But in the final story, Josh, we have lost our Messiah. Nikita Beer is out. Nikita Beer is gone. This is really sad. For those who don't know, he is an internet product legend. He just builds viral products over and over and over again and continues to sell them. he got a role at X as the head of product there. He is in charge of all of the decisions around what the product looks like,

how it feels, how it functions. He did, I think what everyone would call a remarkable job and nothing short of it in turning around the X platform and really improving a lot of the feature sets, making the timeline experience much more pleasurable and being transparent about how they're doing things along the way. He has unfortunately left his role, which makes me sad.

As someone who spends his entire day on X, I really liked having Nikita at the head because you kind of trusted the idea that the product was going to continue to get better, and you would see all those updates along the way and be able to actually participate in the development of what that looks like. He is stepping down. One has to imagine that over, what did you say, the last 400 days, it's probably been nothing short of hell every single day.

Like working at a company that is actively burning down with a very low amount of talent helping you, because we know how lean they are. They've hired a lot of people early on. It must have been nothing short of chaotic. But just wanted to give him a shout out, say, like, thank you. Congratulations. Appreciate your service, sir. And yeah, hopefully the company is in good hands. Elon, it sounds like they left on good terms.

Elon left him like a little hard emoji as a reply to one of these. That's a lot coming from Elon, by the way. Yes, it is. It's a lot. So clearly, Nikita did a good job across the board, but just wanted to pay him flowers and conclude the week on that positive note. But that's everything. We are totally caught up. That's all the stuff.

If you've made it this far, thank you for listening. you can go out for the weekend, enjoy it, go touch some grass, go like disconnect in this sweet, sweet moment of singularity as everything continues to go vertical, as memory sells out, as we get recursive self-improvement, models breaking out of sandboxes. It's chaotic, but it doesn't have to be. You know why? Because most of the people walking down the street have no clue any of this is going on. So not only are you in the know, but you have an edge as to where the future is going. Thanks for watching. Ejas, any final parting thoughts before? Yeah, yeah, I had some homework. I had some homework for the listeners.

If you didn't listen to any of our earlier episodes this week, you should give it a go. We had some ban on our episodes. We broke down the whole Citadel, Leopold Ashenbrenner situation. And we also spoke about the recent SpaceX unlock, which is probably happening. It's probably a day after, and if you're listening to this right now. As we record, it's happening. And how that's going to affect market dynamics and SpaceX trajectory in future coming up. So give those a list and let us know.

Also, if you on subscribe to us, please do. It helps us out massively. Turn on notifications if you're on YouTube. leave us a comment. We read all of them. DMS on X. Heck, I don't know. Just get in touch with us. Let us know if you agree or disagree. And yeah, thank you so much for listening. And we will see you on the next one.