It's already in the public domain (thanks to the OpenAI trial) that Grok distilled OpenAIs models. Listening to the data going into the models in the data centre would be very similar thing. There's some downsides (you're passively listening, not controlling the queries), and some upsides (way more data). But it only ever gets you to some percentage of the existing production model. It doesn't get you what Musk wants - an AI company capable of designing and deploying leading edge models. It gets you to fast follower status.
Seeing both OpenAI and SpaceX trying to yeet themselves into public markets as fast as possible I've got to wonder if the music is slowing. There's a neat rocket company with a satellite business in here somewhere, but it's massively over-shadowed by totally underwater social media site and a failed AI experiment, both of which have been bailed out. I think the best thing you can say is that atleast we aren't aware of any other massively unprofitable bad bets Musk has made that he's going to need to bail out soon.
In some ways this looks like Meta. Meta throws off a tonne of money with it's ad business, but you have to discount it because Zuck has control and an attitude that it's his toy. So you have to discount the ad revenue business because there's a good chance that Zuck just pisses it up the wall. The difference here is you've got a speculative idea that SpaceX might eventually become a massive revenue driver, but Musk is already pissing the money up the wall.
I get why this makes sense for Musk - get SpaceX public, use the stock to merge with Tesla, it gets all his companies under 1 roof and gives him enough voting rights to do whatever the hell he likes. It makes sense for SpaceX early shareholers - they need liquidity.
I do not know what sense it makes for any investor. The absolute best you can argue is it's going to be a meme stock.
I guarantee the reason that number is so high is so that SpaceX can offer to merge with Tesla and the end results will be Musk having over 50% control in the combined entity. He's been very clear that's what he's been attempting to do with via his compensation packages at Tesla and now he's found a different way of acheiving the same thing.
I don't think this is right. Anthropic's growth in the last 6 months went hockey stick in quite an unexpected way (eclipsing OpenAI), so they've done what is sensible - they've increased their compute spend. I don't know if what they're buying from SpaceX is good value, I think there's plenty of reasons to think they got a fine deal. X AI failed. Everyone left. So SpaceX is sitting with a bunch of empty server farms. Yes, Anthropic are desparate for compute, but SpaceX are desparate to IPO a company with double digit billion dollar revenue for $2T so I think there's good reason that this deal represents reasonable value.
The more I read about how big businesses operate the more I think it resembles the weather. There's no intelligence in there, it's just random fluctuations. FiveThirtyEight never made any sense at Disney and seems to have been passed around there more like a trinket than a decades work of dozens of people.
One of the most frustrating things about getting older — besides all the fun stuff that happens to your knees and hair — is the fact that younger generations just take what has been normal their whole lives and say “yes this is the normal state of affairs.”
We used to have laws and limits regarding media ownership. One company couldn’t own every radio station in most of America. Distributors couldn’t own studios. Etc.
Disney should never have been allowed to buy 538 in the first place. ABC, possibly…? But Disney shouldn't be allowed to own ABC!! (And if you’re left-leaning, you can’t pin this mess on the “corporation-friendly” Republican Party because it was Bill Clinton who put his signature on this mess!)
The state we’re in is not normal and it wasn’t necessary and we don’t have to just live with it if we don’t want to.
> (And if you’re left-leaning, you can’t pin this mess on the “corporation-friendly” Republican Party because it was Bill Clinton who put his signature on this mess!)
I would argue Clinton's presidency moved the Democrats from a center-left party to a center-right party, given its platform of welfare reform and free trade at the expense of labor rights and the social safety net.
Yeah, we've only had 3 dem presidents in 30 years and they all collectively moved the dial a certain way. That's an entire generational upbringing of leadership that's a dim shadow of what Carter, FDR, and the ghost of JFK managed to do for its people.
Neoliberal ideals are really good at keeping a machine running in the direction it's already going. And we arguably needed that with Biden for a short spell. It's never going to change the direction of the machine, though. And we're heading off a cliff
Karpathy is probably one of the biggest names in AI, I do wonder where he fits now. He's sort of bounced around Tesla back to OpenAI back to independent. He sort of left OpenAI before it really hit the inflection point, and he was at Tesla for a long time and they didn't really deliver what they wanted on the AI side. Now he's bounced around a few places. I understand that the leaders in this market play this silly game of trying to buy up the names like trading cards but I wonder what this turns into.
i wouldn't be surprised if he just becomes a glorified marketer for anthro.
im also going to guess that whatever research he does would be free roam research that primarily serves to market the fact that claude was able to help perform the research.
the visible stuff he's been working on has been mostly agent soft skills. off the top of my head is autoresearch and his the wiki knowledge stuff. nothing particularly groundbreaking, but has helped devs expand their understanding of the utility that these models can provide.
I think you are underestimating both the value of both projects (autoresearch and personal wiki) just because they are simple. I see both POCs for continuous learning / optmization on the harness layer, which in my opinion is a very interesting direction.
I think Andrej has the experience (and now ressources) to productionize this research into something very interesting.
p.s. called it
> Karpathy will help launch a new team focused on using Claude itself to accelerate pretraining research — an increasingly important frontier as AI companies race to automate parts of AI development.
(https://www.axios.com/2026/05/19/anthropic-openai-karpathy-a...)
No, these are developed off of the assumed uses of the models (predictive autofiller) rather than their actual, cognitive and potential industrial use (developing large scale frameworks for industrial production, automating systems that normally require human monitoring), and uses that we have not yet discovered, because we have not figured out all the constraints and limitations of these models. If Karpathy was in the game like he used to be, he would be on real product. Right now he’s probably so lost by the very thing he helped create that he is stuck doing these mini projects for his own personal interest, without anyone really critically engaging with his work.
Those projects are a complete joke. Neither of them were even original, people have been playing around with those ideas for well over a year.
They just became "famous" because Karpathy is effectively an AI celebrity, so he could throw shit at a wall and post it on X and it would get 10k Github stars.
But seriously, people have been using the models to tweak hyperparemeters, or using LLMs to help create a second brain using markdown or json files or 100X other combinations of files, for a long time already.
yes stop kidding yourself that he is going in as a tech leader in terms of providing technical innovation..at that stage its your persona that matters not the tech (sure I think Anthropic is going to listen to his advice..but its a transactional marketing win primarly)
his value to Anthropic is his influence..he has over 2 million followers, and value is that he is the Top influencer for AI right now, like it or not. just like Selena Gomez might be for top for women age 21-29...
Every AI nerd I know reposts his (very thoughtful posts and projects mind you) like religon
Generally, when a "good" developer has a huge public presence and reputation, that's quite valuable to a company when they're competing in a tough space. Many a time, more so than the (very high) technical skill of the developer in question.
I've seen large funded companies gather good popular developers like pokemon cards and just have them go around give talks and write blog posts. It creates an aura around them which makes things like hiring, fund raising etc. much easier.
So, it's not really a statement about Karpathy himself. It's more about the company hiring him.
Yea, I say this as a marketing agency owner, not a developer or AI researcher, that besides Sam Altman, Dario, Demis and Elon, that Karpathy is one of the most influential I follow.
There’s a lot of value for the business world in learning AI from someone who has been at the top of their game but now is doing a general service by being a great educator and translator between the fields.
His recent Wiki approach may be simple to devs but is certainly an aha moment for the rest of the peanut gallery paying attention!
His LLM-wiki framework has been very useful for me for some personal research and knowledge-building projects I've been working on recently. When I get an idea for a new project, I first give it to Claude together with LLM-wiki.md and have it spend a few sessions compiling knowledge in the wiki before beginning work on the project itself. I schedule further wiki-maintenance sessions for later, too. Over time, the wikis become especially valuable when planning major changes or additions to the projects, as they help to ground both me and Claude with knowledge specific to the project.
Here's an example wiki in a public repository for a dictionary I have been having Claude build for the past few months:
> That implies Karpathy is either dumb or desperate
This kind of thing happens to big names in software all the time. Carmack going to Facebook is a prime example - he joined with the idea of using all those resources to build world-changing tech, and instead he ended up headlining conferences, and fighting a losing battle against the corporate types who were put in charge of Oculus.
Hasn't Carmack solved a few serious engineering problems, making Oculus more or less the most advanced VR device? (The fact that an advanced VR device does not seem to be needed by the mass market is not an engineering problem.)
Yes - but - ironically - he did that _before_ joining them. IIRC he literally started collaborating and helping them while being at a different company.
That seems surprisingly common to me. Visionary engineer has solution to problem, gets hired, solves the broad strokes in the first year, then spends N more years in meetings with exec stakeholders and worrying about schedules/hiring/financials instead of _doing the vision work_.
yeah, and its kind of just misaligned incentives. Visionary engineer wants to solve hard tech problems, corporation wants a product with mass-market appeal. To hit mass-market appeal, corporation cost-cuts until the hard tech problems are outside the solution space...
No it doesn’t? It matches his skills to the lab’s needs. Karpathy is a media personality, manager, and educator far more than he is a hands-on researcher.
It’s kind of useless to argue through metaphors here. There are a hundred researchers with more significant contributions to theory and practice than Karpathy. If you disagree, I’d love to see what papers or implementations you think he’s offered that pushed SOTA.
Andrej is a smart guy. You don't get into Stanford for grad school without that.
But he has always been known for his communication rather than his research. He got famous by putting out a (very well made) course on machine learning that was available to the public. Since graduating he hasn't exactly delivered on revolutionary new stuff at the businesses that employed him but he has continued to be extremely good at communicating thoughts about the current and future state of AI. Businesses want that and he knows that he can deliver that.
I don’t think it does. I think it’s better phrased that he is marketing rather than a marketer. He can do whatever he wants to do, in return Anthropic gets to say “hey, this guy works with us!”
He already stated his motivation a few months ago in an interview with Dwarkesh - basically saying that he might join one of the big labs, for a while, to keep in touch with frontier research.
Andrej seems like a great guy, but him joining Anthropic feels a bit like a transactional relationship (rich old guy marries hot young chick). Anthropic get a "glorified marketer", and he gets a front row seat at SOTA LLM dev 2026. I don't think they hired him expecting he's going to change the direction/pace of their research.
Maybe poor choice of words on my part - what I meant was that this doesn't appear to be a case of AI research co. hires AI researcher to do AI research.
A regular marriage is transactional to some extent too right, but not quite the same as Anna Nicole Smith marrying a 90yr old.
As an aside, an Indian guy I used to work with once explained to me how traditional Indian arranged marriages, like his own, work, and they are HIGHLY transactional. It's not just a matter of same caste, same social status etc, but an explicit trade off. In my co-worker's case he cheerfully told me how his wife was very dark skinned, therefore considered not that attractive/desirable (to other Indians!), but her family had money and social status so it was considered a fair trade for a nice looking boy like himself!
I don't know anything about this person, but want to point out that renown and validation is something that most (all?) humans crave. That doesn't make them dumb or desperate, it makes them normal.
Yes, that's probably his dumbest public idea to date. Given that this GPT repos and parts of autoresearch are brilliant I'm sympathetic. I think he's earned the right to exhibit mild expressions of AI psychosis at this point.
And, my objection was that he clearly had no understanding of the supply-chain risk he was worsening by advocating widespread use of Obsidian for agentic engineering tasks.
Since his announcement, Obsidian has taken proactive steps to mitigate the risks, or at least study threat model. Hopefully, they will implement proper RBAC or something before someone else with his visibility announces an even more irresponsible half-baked idea.
oh my, i see what youre saying. at this point youd hope everyone has realized that the best way to keep models more reliable is to force them to stay honest via very very string static typing as a feedback loop. bags of text with hyperlinks certainly fail that measure
There are things that you can only explore and learn in those places, for obvious reasons.
I don't know his personal life goals but he's a great communicator and educator, if this decision makes him more up to date, and allows him to create even more relevant content then is something everyone will benefit. I understand the risks of being bias toward one company and not the other, but if you look at the content he created so far, he always talk principle first and specific tool later.
I think people here should give him the benefit of the doubt.
Alright, is just that your previous comments and some others sounded to me a bit too judgy, I had to re read these with a new interpretation:
> i wouldn't be surprised if he just becomes a glorified marketer for anthro.
> im also going to guess that whatever research he does ...
At first sounded like unnecessarily prejudging a person and his future efforts.
But if you say you meant everything out of respect for him, then I have to re read what you said , more as skepticism towards Anthropic than picking up on him at a personal level.
I guess what's behind is the sentiment that big companies are going to (or have already) cross ethical lines when it comes to survive and making revenue, and I can share the same concern easily. I think in this forum the majority would do.
I would just make that more explicit and separate the person from his choice, and his future work from how the company could use it, that would be a more respectful way in my view.
But is just my opinion! and I'm aware I'm might be picking on nuances that don't matter or being overly polite :D
he's not dumb or desperate compared to the average person, but it's very possible to be dumb and desperate compared to the delusional promises and outsized amounts of money in the industry. Manages to make smart people look extremely stupid every day.
He may not be a brilliant researcher, but he is a brilliant teacher. I am glad he is joining Anthropic so he can stay up to date with the next round of things that he will teach :)
> It is a pity we don't hear more about the truly brilliant researchers.
Reason? What is the value of that other than entertainment? And it's not in the interest of companies to make celebrities that then are poach targets (if they can avoid they would yes there are exceptions as noted elsewhere in this thread).
And if you did 'hear' (via articles) to what extent what was said even be correct vs. a writer just fluffing things up to the max.
Tech is not sports where you can actually see the superlatives and know that the person who praise is being lavished on actually won or threw or caught and so on. (Or even music where you can hear it and see the stadium that is packed with fans..)
Tesla self driving kind of works. In a very similar way to how it kind of worked back in 2016. It's better than it was in 2016, don't get me wrong. But even today they haven't solved the problem and Karpathy left in 2022. And other companies notably have actually surpassed Tesla over that time. I don't think anyone could reasonably say he walked away in 2022 because he thought the job was finished.
When OpenAI was founded, the mission was to develop AI, but nobody (anywhere) knew how to do AI, so OpenAI did ML research on games instead, which is what DeepMind was doing (with Google's perceived AI/ML dominance being the raison d'etre for OpenAI, and Google having just bought DeepMind). This was the era when Karpathy was at OpenAI.
Around the time Karpathy left, Ilya Sutskever, another OpenAI founder, started playing with Google's new "Transformer" architecture, which was the beginning of the "GPT" series, GPT-1, GPT-2 and eventually ChatGPT (GPT 3.5 + RLHF). In retrospect OpenAI's early Transformer experiments and GPT-1 was the inflection point that moved OpenAI from a company that wanted to build AI, as soon as anyone else did, to one that was actually doing so, although I think it would be revisionist for anyone to claim they knew what they were doing at the time. The early GPT-1 and GPT-2 papers read more like "wow, this is a bit unexpected, look at all of the things it can do!".
Well Karpathy left in 2017, and all the sort of commercial stuff didn't happen till a while later - for example they set up the structure to take external money in 2019 and that's obviously the point at which they'd found the pathway that justified doing massive training runs and all that. So Karpathy was out very early (left at the point that Musk thought OpenAI had basically failed).
FWIW while Karpathy was at Tesla he was basically working on the vision component. The actual driving component (using vision as an input) was originally all C++. They may have started migrating parts of the driving component from C++ to neural networks while Karpathy was there, but most of it happened after he left in 2022, with the big switch being FSD 12 in 2024. User reports from before/after FSD 12 are like night and day.
I'm not talking about vision (cameras) vs lidar etc, just the Tesla FSD architecture that separates the "vision" component (turning camera/sensor inputs into symbolic road/sign/vehicle/pedestrian/etc data), and the driving component which takes the vision data, plus current location and destination, and uses that to actually drive the car - switch lanes, make turns, avoid obstacles etc etc.
Some people are good at developing the sciences. Others are good at developing commercial products.
And tesla is not a good place for science development. Tesla is structured from narcissistic mindset: results driven, cynical, and position-based. This doesn’t bode well for long term sciences.
I read this as a bad sign for Anthropic. Relying yet again on more hype instead of improving products.
OpenAI’s hiring recently has been much stronger, whether through luck or structure. The “no-name” guys have actual taste. I love that. I don’t care that they’re no-names.
I don’t know Karpathy personally, I won’t speak bad about a man I don’t know. I hope he makes CC better. I just read this as hype. My point is that there’s nothing he has that an empowered no-name product manager doesn’t. It’s like Alex Wang at Meta. That acq didn’t redeem Meta. They actually lost LeCun. Where’s Llama today?
Regardless of what Anthropic’s share price is, OpenAI has been fucking killing it recently. I don’t take particular pleasure in saying that, i’ve been a google and gemini guy for years
My lens is meritocratic. My experience is as an extremely heavy user of both company’s full suite of products in the range of 5 digits per month. My interest is better products not hype.
>OpenAI’s hiring recently has been much stronger, the guys have actual taste.
Can you cite specifics? "I won't speak bad about someone, but also won't speak good about others" resulted in a comment that seems to contribute nothing
I’m hoping Karpathy will make Claude Code better, in the meantime I’m super happy seeing a small product manager like Tibo fucking crushing it on Codex
Joking aside, there are small communities pushing codex and AI to the bleeding edge of what's possible.
Here I'll give you an example. The last few updates from Boris at CC have been tweaks to the system prompt to make it use less compute, effectively making the system dumber, making it tell you to go to bed. I mean come on! Tibo has been impressing me, bc they're building the things these small communities are building.
One of the things these bleeding edge guys and girls have been working on is a /goal feature, essentially ralph loops. Codex released it as a feature the other day. I can't help but be impressed. As an ex-pm, this is product management.
Then you take a look at what the Chinese are doing on their own forums, and it just makes what Google and Anthropic are doing look outdated. OpenAI feels competitive, which I like. What's coming will not be kind to us, we adapt or we die.
I really hope the Chinese win this, the Epstein class deserves everything coming for them. Btw, the Chinese president is much more moral than any of the creeps leading USA currently, just see the wikileaks cable assesment on Xi Jinping and why he was not part of the Epstein class.
It feels like these companies are constantly going back and forth on who has the best product constantly. It's such a dynamic time with how fast they are both working.
To be fair, Mythos is probably one of the most significant marketing pushes in the industry in both impact and investment.
I am sure there is an element of reality in it's capabilities, but there's also a significant amount of "We don't have the compute to handle this at scale", and "look look, we have the best model. It's so good that you can't even compare it to other models. That is how good we are."
I’m noticing a real disconnect in the user base about this
The Claude maximalists that can never see any wrong in anything and the users that care about actual capability
These guys are going to be in for a rude awakening when the Chinese are steamrolling us with data centers you can see from space and better models, Amodei will tell you that himself
Hey, it's not like the Chinese have a serious demographic crisis they can't cope with, and their only hope is to significantly increase productivity per worker.
Whenever I see a user base turn against actual users or imply censorship or discredit actual experiences it always ends in a death spiral: Deny -> Product stops improving -> Censor -> Die
Dude, both! Codex is going to eat Openclaw… i don’t love saying that.
What codex is a few steps away from doing is changing fundamentally a lot of workflows.
Remote codex with their computer use is basically you at your computer doing things, 24/7.
Then they added gpt images 2.0
what codex can do, in a few more product iterations, is show you visually side by side “would you prefer this (A) or that (B)” in a series of questions. This is what some open source researchers have been up to. That’s no longer guessing.
I’m not trying to hype a company i have no stake in, but they’ve been killing it.
It’s extremely compute intensive, but also very satisfying.
Out here in the actual demonstrated world, OpenAI has been leaking quality people like a sieve, has not yet demonstrated anything remotely similar to 'taste', and is led by a sociopath (https://www.newyorker.com/magazine/2026/04/13/sam-altman-may...), so I think you can rest easy.
I mean, you would think that all those people he killed as the person in charge of deploying knowingly dangerously defective self-driving software for profit would have had a impact. But executives seem to just skate on killing customers to line their own pockets these days. Just "following orders" I guess.
Yes, he was [1] director of AI and Autopilot Vision at Tesla, directly poached and reporting to Elon Musk on the most important headline feature of Tesla directly managed by Elon Musk.
He had both the technical and executive authority to determine if the product was fit for customer usage. He had direct executive responsibility for the product on the road between 2017-2022.
If he, the lead architect and executive responsible felt the product was dangerous and then he was overridden, he can not get away with claiming he was “just following orders”, he had a moral duty to not sign-off or quit otherwise he is clearly complicit in deploying a dangerous product for his own self-enrichment.
When people talk about engineering ethics, this is literally a completely uncontroversial textbook example. The only way you accept this is if you do not want ethics in engineering.
Furthermore, he was extremely hireable with numerous job opportunitys available to him. He would not be destitute or even particularly worse off if he did quit for ethical reasons. Any self-preservation defense is also invalid.
Andrej Karpathy is a reason* Tesla doesn’t have Lidar and thus is a reason Tesla self driving isn’t nearly as safe as it could be?
He heard Elon say “I drive with eyes, so cars just need eyes” & shipped?
:( happy to have my impressions corrected (but I was kind of pretending it’s a 2026 scenario where you could slap Lidar, ship a Waymo, if you were just willing to spend the friggin MONEY - 2017 was too early for most any “self” driving IIRC)
-
*edit - in a scenario where his refusal to skip Lidar catalyzed change
According to ABET they do if they want the degree to be accredited. We had two classes for my SE degree. From Criterion 3. Student Outcomes:
"2. an ability to apply engineering design to produce solutions that meet specified needs with consideration of public health, safety, and welfare, as well as global, cultural, social, environmental, and economic factors."
"4. an ability to recognize ethical and professional responsibilities in engineering situations and make informed judgments, which must consider the impact of engineering solutions in global, economic, environmental, and societal contexts."
Passing a mandatory class != believing in its message and acting on it.
Unfortunately, rather important courses like engineering ethics have become lumped in with mandatory DEI objectives and similar 'grievance studies' requirements, classes which many suffer through quietly, regurgitating the Correct responses while they count the minutes until they can get back to more substantive classwork. Some undergraduates may unfortunately gloss over ethics just as they gloss over lectures on privilege.
I somehow felt he, along with Andrew Ng, are very few well-known AI experts that are left behind on the money side during the AI-gets-me-super-rich crazy time, unfortunately.
I can't speak for Andrew Ng - but my take is he did out of pure altruism - love. just in terms of advancing free education e.g coursera & the free machine learning courses etc he brought to the masses.
His goal could simply be to learn SOTA architectures.
When rumors started that GPT-4 design would be kept secret, he likely wanted to know what architecture it would be. Perhaps he left Tesla, waited out the non-compete clause, and joined OpenAI to learn its details.
When Mythos dropped, there were hints that it had a new architecture. He might similarly want to know how it works.
Either way, there is enough cross-lab hiring that those secrets eventually get known, but only by the labs.
Is this the right view? I would imagine Elon Musk has essentially infinite demand for lawyers. The choice for the lawyers is probably work on this absolutely dead case where all the facts are not just against them, but highly embarrassing for their principal, or work on some other case where there would be perks like having the possibility of winning the case. I'm quite certain highly trained, highly successful, expensive lawyers aren't jumping up and down in excitement to work on cases they quite well know are meritless.
They'll do it, because that's what Elon Musk wants them to do and he's paying them and they want his other business, but this isn't what gets them up in the morning.
I think it’s just that you can make an entire career (and a shitload of money) by simply being in Musk’s (and other rich people’s) orbit. From handling his lawsuits all the way down to cooking his food. He is an infinite River of money and you just dip your spoon in a little.
Can we just bring some reality to this conversation. The non-profit still exists, it owns a significant chunk of a now almost trillion dollar company, and the for-profit wouldn't be profitable if it weren't for Billions of dollars of investment from Microsoft.
There is no counter factual where OpenAI exists as a non-profit and still inexplicably gets handed billions of dollars of compute to train LLMs. The for-profit company is a different thing from the non-profit and it exists for perfectly understandable reasons and I'm unsure why anyone other than Elon Musk pretends this doesn't make sense.
What is your proposed counter factual where the non-profit entity retains all ownership of the venture, but somehow finds hundreds of billions of dollars to train LLMs?
An alternative chain of events might be one where OpenAI, Inc. only operated through income and donations, which would probably lead to them scaling more slowly. They could also take on loans.
Having thought about this whole thing more I believe that non-profits should not be a separate type of entity that gets special tax treatment. People have different ideas of what constitutes legitimate non-profit activity. It just opens things up for tax avoidance and scams.
reply