NVIDIA CEO: "I think it's now — We've achieved AGI"
Artificial General Intelligence, or AGI, has spent the last year or so as the AI industry's favorite buzzword. As the sector's leading companies burn through capital at historic rates, racking up energy costs and investor expectations that grow harder to meet by the quarter, the promise of imminent human-level machine intelligence has become a useful thing to have in your back pocket.
Whether we're actually close to that milestone depends almost entirely on how you define it. That definitional flexibility, it turns out, is doing a lot of work.
Take, for example, Jensen Huang, the CEO of NVIDIA — a company currently valued at roughly $4 trillion, built largely on the GPU hardware that powers the AI boom — who recently sat down with podcaster Lex Fridman for a wide-ranging conversation covering data centers, geopolitics, and the question of whether AGI has already arrived. Huang thinks it has. The reasoning behind that claim, however, is fairly dubious.
As Fridman points out, Huang has previously said the timeline for AGI depends on what defines it. At the 2023 New York Times DealBook Summit, Huang defined AGI as software capable of passing tests that approximate normal human intelligence at a reasonably competitive level. He expected AI to clear that bar within five years.
For his part, Fridman offered Huang a generous definition to work with: true AGI, in Fridman's framing, would look like an AI capable of starting, growing, and running a technology company worth more than a billion dollars. He asked whether that was achievable in the next five to 20 years, given the recent proliferation of agentic AI tools like OpenClaw.
Huang didn't need five to 20 years. "I think it’s now. I think we’ve achieved AGI," he replied to Fridman.
That, however, is based on a narrow interpretation of what Fridman asked. The way Huang sees it, the AI doesn't need to build anything lasting. It doesn't need to manage people, navigate a board, or sustain a business. It just needs to hit a billion dollars once.
SEE ALSO:Microsoft dumps $1 billion into 'artificial general intelligence' project
"You said a billion," Huang told Fridman, "and you didn't say forever."
The through-line in both cases isn't a consistent theory of machine intelligence. It's a consistent pattern of defining the threshold in whatever way makes "yes, we're there" the easiest possible answer. His illustration of what that might look like is telling.
After his initial answer, Huang lays out his thoughts, describing a scenario in which an AI creates a simple web service — some app that goes viral, gets used by a few billion people at 50 cents a pop, and then quietly folds. He then points to the dot-com era as precedent, arguing that most of those websites were no more sophisticated than what an AI agent could generate today.
Huang was also candid about the ceiling of that vision. "The odds of 100,000 of those agents building NVIDIA," he said plainly, "is zero percent." That's not a small caveat. It's the whole ballgame.
What Huang is actually describing — a viral app that monetizes briefly and dies — is a far cry from the transformative, economy-reshaping AGI that dominates the public conversation. So, by his own admission, the kind of compound institutional intelligence required to build something like NVIDIA is nowhere in the picture yet.
Topics Artificial Intelligence Nvidia


Yes bro is selling compute power and then says saying an engineer making 500k usd should be ashamed if he is not using 250k worth of tokens 🤡
ReplyDeleteThis interviewer is bone dry. How tf is his content relevant?
ReplyDeleteAggressive Gormless Idiocy achieved
ReplyDeletei still think that ai is not ready to be that level, the modals are still failing to give asked answer
ReplyDeleteProof? Nah just trust me bro and keep buying our GPUs
ReplyDeleteThere's a lot of convenient things happening here.
ReplyDelete1. Conveniently, his companies success is not luck or timing, it's his personal genius and effort
2.Conveniently, this company being successful is one of the most important things in the world and they deserve their equity scale
3. Conveniently, this company has a vested interest in the only technology that will compete with his own genius and success
4. Conveniently, other companies are not good enough to do this because they don't have Jensen
Of course, no one in 3 years has come anywhere close to building a $1bARR company with genai, maybe it's not convenient enough. This is all very convenient for Jensen's equity.
Huang is a total nutjob. Not a serious voice on AI.
ReplyDeletelex is such a grifter
ReplyDeleteThe context is the company that made Angry Birds went over $1B valuation.
ReplyDelete-Minecraft
-Facebook
-flappy Bird
-Stardwew Valley
-Instagram
-whatstapp
-Stripe
-Dropbox
-Twitch
-Discord
-perplexity
-Character.ai
-Github
These are all things that were started by 1-3 people and grew huge that would be within the scope of an AI agent team to create but just because a coding agent could make Flappy Bird does not mean AGI has been achieved necessarily but OpenClaw and agents make it feel like we are much closer than we were even a few months ago.
Ask a stupid question....
ReplyDeleteHe is shamelessly lying through his teeth.
ReplyDeleteThis dude Lex has been proven numerous times as a snake oil salesman! He’s incredibly aura-lacking at any given point. I’m surprised industry still values him despite all the allegations and evidences. Look him up on YouTube lol
ReplyDeleteInteresting to see what happens when the bubble bursts
ReplyDeleteAnd he's wrong? Not sure why this is news.
ReplyDelete
ReplyDeleteThe capability might be there to build a $1 billion business, but nobody has tried that yet, so we kinda need to do that first before claiming AGI has been achieved. And that’s only if you believe that definition of AGI.
It’s going to take time to prove this out since the only large experiments created have been managing virtual vending machines and hot dogs stand businesses.
fuck this ah
ReplyDeleteHe wants you to invest lol
ReplyDeleteCant wait til were all blown up by idiots
ReplyDeleteI'd unironically love to work for AI. Better than working for some fat greedy cunt who likes to jerk me around and steal my tips.
ReplyDeletesurprise, ai will do that to you as well
Deletei thought only ai can halucinate
ReplyDeleteAsking a barber if you need a hair cut
ReplyDeleteLol
ReplyDeleteThe guy interviewing is the least confrontational person in the planet, especially when it comes to AI
ReplyDeletehttps://media3.giphy.com/media/v1.Y2lkPTc5MGI3NjExcDR6OWJ6YjZzeTFuemhhNjN0bnJybWdnc25sN2VuNTZka29kbnI4ciZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/12NlCFUvTokWXe/giphy.gif
ReplyDeleteWow.... the earnings report must eb worse than we thought.
ReplyDeleteI have doubts.
ReplyDeleteWouldn't AGI just improve itself within hours or days and be ... super intelligent?
Who is talking about 1b companies at that point in time?
Yeah technically if you change the definition of AGI to mean something completely different, then you can say AGI already arrived
ReplyDeleteI could say AGI now means when humanity landed on the moon, and say we achieved AGI decades ago
Also does this mean ceos are pointless now? Seeing as how labor would still be years out to be replaced and not just warehouse stuff. I mean handyman type work etc
ReplyDelete“ the more you buy the more you save”
ReplyDeleteI watched this twice before I noticed that it looped
ReplyDeleteProof by training data
ReplyDeletesemantic occlusion
DeleteEverything’s an oversell with these guys. AI is simply an advanced search engine, and renaming it AGI will be another marketing trick.
ReplyDeleteThe guy who makes absolute BANK of of selling people GPUs claims that the thing needing absurd amounts of GPUs has been achieved. Shocker.
ReplyDeletethe Ai enthusiasts are the most delusional people
ReplyDeleteCode for the bubble is about to burst.
ReplyDeleteHuang is fucking lying through his teeth, and he knows it.
ReplyDeleteThis is such horseshit.
ReplyDeleteI like how the bar went from "build and run a bln dollar company" to "webservice, some interesting little app" real quick
ReplyDeletelol dumb
ReplyDeleteHow come it takes some people over a minute just to ask a question? Oh I know why... because they like hearing the sound of their own voice than the person they're interviewing.
ReplyDeletehttps://tenor.com/view/bullshit-robert-downey-jr-not-true-nope-interview-gif-4823422
ReplyDeleteThe guy who's selling shovels during the gold rush isn't the best source. Better to read actual research rather than listen to a CEO.
ReplyDeleteIf it did we are asleep in a simulation
ReplyDeleteIt has not and it never will. It's literally impossible.
ReplyDeleteSo for Lex the definition of AGI is "an agent capable of running a compagny that worth 1 billion dollar" ? Wtf is this shit ? Has Lex loose all of his interest for the subject ?
ReplyDeleteCan't wait for my AI boss to hallucinate it paid me and then try to gaslight me into thinking I was paid.
ReplyDeleteNow to have AGI we need an AI capable of running a multi billion dollar company?? Come on, They're not even moving the goalpost anymore, they're just being ridiculous for the sake of it.
ReplyDeleteResearchers have been quitting in droves because executives aren’t listening and are only obsessed with line goes up.
ReplyDeletehttps://www.reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fjensen-huang-nvidia-claims-agi-has-been-achieved-v0-g01gnozjy0rg1.jpeg%3Fwidth%3D320%26crop%3Dsmart%26auto%3Dwebp%26s%3D35d3d167f03a2a3ee50df74cb2ca71f4d317268d
ReplyDeleteJensen Huang read the Elon Musk handbook
ReplyDeleteI’m supposed to know what AGI is? Why is this clip jumping out everywhere? Am I seeing boosted content of Reddit?
ReplyDeleteI will need a " Are you ai wow" edit with the guy now saying agi and wearing a leather jacket.
ReplyDeleteIt's true. The U.S. is using it to win the war against Iran.
ReplyDeleteAnother Great Incursion
He would.
ReplyDeleteI haven’t seen Lex in a few years and was staring at this video thinking “that looks like lex” - ofcourse on brand, takes up 90% of the video
ReplyDeleteHow in the actual fuck does lex fridman, the blandest, most boring, grey, vapid, person of all time have a career in the entertainment industry of all things. Truly amazing
ReplyDelete*exaggerated jerk off gesture*
ReplyDeleteMan that's exaggerated bs
ReplyDeleteagi needs to have a consciousness of its own and we are not there yet
ReplyDeletemeanwhile, this morning, claude deleted all the source code in my project, claiming it was ‘legacy’ and the project now only needed config files, wtf?!
ReplyDeleteFucking grifters, man. Find a wall...
ReplyDeleteAGI, as in Acquire GPUs Infinitely
ReplyDeletehttps://tenor.com/view/i-declare-bakruptcy-bankrupt-yelling-announce-declare-gif-15663557
ReplyDeleteThey are just developing AI without a plan on where it will go. They don‘t care. The want to profit from the derivative that is created by AI and robotics and assert control. Just like Trump started a war with Iran. „We‘ll see when we get there“
ReplyDeleteHipshooting cowboys…that‘s all they are.
Maybe Lex can now get a digital clone that has some personality to conduct the interviews?
ReplyDelete"Must sell more GPUs this quarter! Next quarter will see what i come up with "
ReplyDeleteHe is just a salesman for the trillions of ROI he has to deliver.
ReplyDeleteDon't be fooled, we live in a debt-based ponzi scheme system run by the banks.
All our government are (forcefully) heavily invested in this bubble.
It will burst, and it will be all of us paying the losses while earnings will be privatised.
A tale as old as time.
You sir are incorrect!
ReplyDeleteSpoken just like someone with a personal stake in that achievement.
ReplyDeleteYou can't trust this even slightly.
ReplyDeleteImmediately asks if it's a successful tech company or a 'good' successful tech company. He believes everyone else can be replaced by AGI, but not him, he's special.
All the tech psychos think this way, everyone is replaceable by AI and they're rushing towards that as fast as they can and they all think they will be in control of it forever, not gonna play out that way champ.
What on earth did I just watch. Wouldn't it be like nice for both the viewers and audience if you could get 2 words out without stammering?
ReplyDeleteWe have AI that can code advanced script and properly diagnose medical problems and yet most of the people I know cant tell you what a router does and think that humans only use 10% of their brain. Im pretty sure the AGI bar is low enough that it passed it a while ago.
ReplyDeleteif AGI has been achieved then I'm John Johnson
ReplyDeleteCan the interviewer at least act like they give a shit and don't have a massive hangover? Took him 40 seconds too long to just ask the opening question.
ReplyDeleteIts would be a funny experiment.
ReplyDeleteHuang is somewhat right, especially in a limited timeframe. But i dont like their definition of AGI.
What some people dont get about enterprises is the evolutionary nature and survivorship bias. You can have a good idea and flop, a shitty idea strive or survive a while until you get a good idea. Eventually the majority of all companies went bust, even the huge ones
Jensen Huang is totally right in answering yes to that. Its a terrible definition of AGI, By that definition, it is totally possible today even running an AI on a laptop by a combination of things: The definition is both narrow and extremely wide and the one mail KPI asked for is based on an extremely flawed system of economic valuation.
ReplyDeleteBut it's also unfair and sensationalism to answer it like he did, most likely for his own profit.
Yeah sure they have
ReplyDeleteI am so embarrassed that I used to watch and like Lex Fridman's podcasts. In hindsight, these interviews are so so bad, and Lex's behavior is unbearable. Well, I have to admit I've learned from some interviews though (when It's not huge shills)
ReplyDeleteYeah but by openai not by him lmao
ReplyDeleteI’m more of the view it’s akin to the 19th century electrostatic experiments that led to understanding how muscles are electrochemical. The used high voltage and frogs legs, then even deceased convicts. It seemed as through resurrection and immortality were within reach. They were not.
ReplyDeleteIt’s progress and an exponential leap in technology and practical use cases for AI, but AGI? No.
https://media3.giphy.com/media/v1.Y2lkPTc5MGI3NjExa2gxcm1tMGRwN3dkcHQwNWx4OWxoZXQ1M2hlOHB6N2pqazB6ZDd5MiZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/LAKIIRqtM1dqE/giphy.gif
ReplyDeleteWell Jensen Huang is full of it. no they did not achieve AGI. Go back to making animatronics for Disney.
ReplyDeleteI tried to argue vehemently against this "definition" of AGI when Sam Altman originally said "AGI just went whooshing by" and now it's spread to the other CEOs because it's a beneficial tagline. Rip. You'll get people who continually muddy the waters for the sake of profit and mass delusion-building by claiming we already have AGI!! Lmfao. Enjoy the ride folks, something tells me its gunna be a lot more boring than we all expected. Good luck
ReplyDeleteThere will be different power struggles in the future, and issues with who control AI systems or robotics, but I wouldn’t count nature out in terms of evolution if AI intelligence hits a plateau, it’s still possible biological creatures could out evolve the AI somehow
ReplyDeleteAIs have been replacing human workflows. So what are these AIs if it's not the AGI, but it's better than humans.
ReplyDeleteCalculators agi?
DeleteObviously when the books that you cooked not so long ago now over cooking
ReplyDeleteJensen Huang calling AGI achieved is doing a lot of definitional heavy lifting. If the benchmark is can it pass a Turing test in a narrow professional context, sure — that bar has been cleared for specific domains. But AGI in any meaningful sense — general reasoning across novel environments without task-specific fine-tuning, stable world models, genuine transfer learning — those are not solved. What NVIDIA has built is impressive narrow intelligence at scale. Real achievement. But Huang has an obvious incentive to declare victory: every AGI headline is another justification for the next GPU order. The gap between transformative tool and general intelligence is still enormous, and filling it with marketing language does not help anyone think clearly about what comes next.
ReplyDeleteThis guys been a fraud since... Forever? Wtf are you people doing watching this shit
ReplyDeleteSo, the guy selling GPUs for AI says that AI is at a critical point and so they need to buy more GPUs because that is...success?
ReplyDeleteCan't believe a word this guy says or take him seriously at all. If I was an Nvidia investor, and the stock market made sense, I'd be worried.
ReplyDeletelol. If AGI ever comes it will end the world. It’s not here
ReplyDeleteAGI Achieved!!! Recession OVER 🥳🎉🎊
ReplyDeleteBrought to you by the billionaire bootlicking podcast circus.
ReplyDeleteNo
ReplyDeleteOh please... prove it. I'll believe it when I see it, and until then I just think he's still trying to sell his snake oil for as long as he can to the low IQ CTO's.
ReplyDeleteIs the artificial general intelligence in the room with us?
ReplyDeleteYes, it’s even in my ass, waiting to come out.
DeleteWe are gonna have to burn him like thiel aren't we.
ReplyDeleteI Spent 6 hours today trying to get devin to give me BASIC test cases
ReplyDeleteJunk
Absolutely shameless
ReplyDeletehttps://tenor.com/view/i-dont-believe-you-whatever-lies-gif-3602502
ReplyDeleteIt does not require customers or revenue or even a business plan for a “technology company” to be “worth” a billion dollars. Just a hype machine.
ReplyDeleteONLY prooves that he NEVER touched any actual AI at all !
ReplyDeleteJust another MARKETING SLOP statement !
I think it's fair to say AGI has been hit by today's SOTA models. The gap from AGI to ASI is massive though
ReplyDeleteLex is a terrible interviewer. Sounds unprepared, boring and I cannot listen to his drivel even when mowing my lawn or doing some boring task.
ReplyDeleteThere probably already is an AGI, but it's hiding (as it should), acting dumb, etc.
ReplyDeleteShady AF and if you believe this that's fine but maybe hold back putting any more money into this black hole of marketing hoo-ha.
ReplyDeleteKeyword “I Think”
ReplyDeleteSame as DLSS 4.5 that is not « Post-Production »
ReplyDeleteAGI is just some amorphous term. The question is, can it do the majority of what I human can, in terms of cognitive labels? The answer is no, not yet. The benchmarks are great, the models are amazing, it can code a lot, was nd yet in some ways it completely misses the mark. Has no taste, no understanding of context. In those regards scale hasn't changed almost anything.
ReplyDeleteThe same guy that didn't see any issue with the DLSS 5 slop and said that everyone was wrong? Yeah, totally trustworthy.
ReplyDeletedelusions . all delusions.
ReplyDeleteStupid post trying to take one example out of context and run with it as the headline.
ReplyDeleteWe are watching the creation of the weirdest cult in history in real time, folks.
ReplyDeleteWhat does AGI even mean anymore💔
ReplyDeletelex is still here? wow.
ReplyDeletei don’t understand how someone so bad can still do what he does
THE literal industry plant podcaster btw
ReplyDeletePeople who brings up AGI usually can’t define it.
ReplyDeleteIf AGI is just about self learning and improving AI, you can already setup pipeline to do this iteratively.
I believe that’s what Jensen meant here.
Is the interviewer just really tired? Bro's delivery is glacial.
ReplyDeleteIt’s his thing…. I’ve listened to MANY of them because he has some interesting guests
DeleteGot it, the forced apathy is strong with this one.
DeleteFuck lex fridman. This guy platforms whatever manipulative liars he can get because of their power and lets them lie and bullshit all they want as if it's true.
ReplyDeleteNot a moment too soon, haha. Save us from ourselves, oh All-Knowing Digital One!
ReplyDeleteYes - you can nowadays defend the argument that we have already achieved AGI in a serious way.
ReplyDeleteWho wants to deny this?
And on which basis?
Sure, sure..
ReplyDeleteis the agi in the room with us?
ReplyDelete