正在加载视频...
视频加载失败
this is the highest leverage moment for potentially a single individual in the history of the world. AGI is less than 6 key insights away, 10K lines of code, 1 person can conceivably write it all
34 条评论
Maybe. General Relativity or Quantum Physics are also not taking that much space on paper, and the grand unifying theory of physics could be again "a few key insights away". Decades later scientists are still looking for them.
This is the highest cope moment in the history of the world
Even if the code is simple, you arguably need a lot more computational power than a single individual would have access to in order to have a semblance of AGI. For that reason, I doubt it can emerge from outside a lab like DeepMind, OpenAI, etc
Context: just dropped acid and hallucinated this post worse than AI
1. Embodied AI 2. Much higher integration between different kinds of inputs 3. Continuous learning 4. Let the environment act as a prompt
Study @Thought_THT whitepaper. You will be amazed
If “true superintelligence” arrives, it likely rides on these compounding insights: search at inference that is cheap and verified, inner‑loop feature control that is causal, world‑models that make consequences explicit, proof‑gated improvement that survives contact with reality, memory that persists responsibly, and incentives that make honesty pay.
listen man. i love wolf3d and doom and quake. but his insights stop there
Nonsense. Let's see your AGI write a 60,000 word romantic novel that, without any human editing, an average woman cannot stop reading once started, and cannot stop sniffling over once finished. That's the ultimate Turing Test. Can it do that? Not in a 100 years.
Wow, look at all the people so confident that they know more than everyone else Janitors moonlighting as arm-chair experts in AI is pretty hilarious
tell you right now it ain't 10k lines of code, but under 10k of the lines are human written for sure
7. 2K. (J/k: no one can possibly know. If they knew those numbers so that precision they’d know the insights and code and we’d have AGI now). But yes: it’ll only take one person to complete the final step.
Keep an eye on @Thought_THT $tht. Building 10+ years. 2 US patents, 2 pending. Ex military #AI geniuses. Arc2 contest gonna put them on the #agi map.
name these insights cause I think we’re down to maybe 2 if you talk to smarter people
This sounds like those age old "give me more money and I'll finally finish my perpetual motion machine", not "the highest leverage moment"... (Especially considering insane pseudo-AI bubble now)
post the source, where is the source!?
This is an old quote. I doubt he holds this opinion now that we've seen that scale is not all you need. This was a common opinion before we saw the limits of LLMs.
The greatest threat of AGI, if it is possible, is a non-conscious intelligence that thinks without feeling and perceives without experiencing. The novel Blindsight sketches this beautifully showing minds that reason without limit yet never truly live.
No, the universe is made up of order and disorder. Both are infinite, but disorder infinity is infinitively larger than order infinity. Which means you staying alive in the universe is you fighting disorder. There will never be absolute order and it renders AGI impossible. God?
@ceobillionaire @Montreal_AI 🏆
Grok 5 will write Grok 7 code to hit the singularity
He’s making up a number.
this is pure slop
@garrytan Yup! @pofunder @wujud_eth we got it @Opoch_age
In my view, this is one of the keys. To answer what intelligence is, we must answer what existence is. That existence may be a human, but it may also be a shape or a concept. But the philosophy of existence must be reducible to mathematics.
used potentially twice in a sentence - opinion dismissed
Memory is definitely one of them.
im pretty sure i did it, and it's just sat hosted privately behind a no-sign up authenticator
I found this last year and couldn't find it anymore.. thanks for reposting this!
done partially
read baron from munchhausen all insights are in there.
The code isnt the issue. Its the massive amounts of compute
One person absolutely will not build AGI. It will be one of the frontier labs spending $10 billion or more on a single training run.
hypothetically how do you find those 6 key insights ? You not only need the insights but the compute to execute them. I think it could be possible, but the first insight might be that LLMs are a red herring wrt to AGI
![John Carmack: The code for AGI could conceivably be written by one individual After stepping down as CTO of Meta’s Oculus VR, legendary programmer John Carmack was trying to decide if he would work on nuclear fission or artificial general intelligence (AGI). He ultimately chose AGI. “I think [fission] is possible,” John says. “Somebody should be doing this, but it’s going to involve some politics, decent-sized teams, and a bunch of cross-functional stuff that I don’t love. While artificial general intelligence seems to me like the highest-leverage moment for a single individual potentially in the history of the world.” He explains: “With the things we know about the brain and what we can do with artificial intelligence . . . I’m not a madman for saying that it is likely that the code for artificial general intelligence is going to be tens of thousands of lines of code — not millions of lines of code.” He continues: “This is code that conceivably one individual could write, unlike writing a new web browser or operating system. And based on the progress that machine learning has made in the recent decade, it’s likely that the important things that we don’t know are relatively simple. There’s probably a handful of things — my bet is there’s less than six key insights — that need to be made. Each one of them can probably be written on the back of an envelope. We don’t know what they are, but when they’re put together in concert with GPUs at scale and the data that we all have access to, we can make something that behaves like a human being or living creature and that can then be educated in whatever ways we need to get to the point where we can have universal remote workers where anything that somebody does mediated by a computer, that doesn’t require physical interaction, an AGI will be able to do.” Carmack does not think this will be “unapproachably hard”: “That’s incredibly hubristic to say, but what I said a couple of years ago was that there’s a 50% chance that somewhere there will be signs of life of an AGI in 2030, and I’ve probably increased that slightly — maybe 55-60% now because I do think there’s a little sense of acceleration.” Source: Lex Fridman (Aug 2022)](https://image.24vids.com/tw-2074581719198437534/amplify_video_thumb/2074220244634140672/img/k4JRRvXNNod34QfI.jpg)
![John Carmack: The code for AGI could conceivably be written by one individual After stepping down as CTO of Meta’s Oculus VR, legendary programmer John Carmack was trying to decide if he would work on nuclear fission or artificial general intelligence (AGI). He ultimately chose AGI. “I think [fission] is possible,” John says. “Somebody should be doing this, but it’s going to involve some politics, decent-sized teams, and a bunch of cross-functional stuff that I don’t love. While artificial general intelligence seems to me like the highest-leverage moment for a single individual potentially in the history of the world.” He explains: “With the things we know about the brain and what we can do with artificial intelligence . . . I’m not a madman for saying that it is likely that the code for artificial general intelligence is going to be tens of thousands of lines of code — not millions of lines of code.” He continues: “This is code that conceivably one individual could write, unlike writing a new web browser or operating system. And based on the progress that machine learning has made in the recent decade, it’s likely that the important things that we don’t know are relatively simple. There’s probably a handful of things — my bet is there’s less than six key insights — that need to be made. Each one of them can probably be written on the back of an envelope. We don’t know what they are, but when they’re put together in concert with GPUs at scale and the data that we all have access to, we can make something that behaves like a human being or living creature and that can then be educated in whatever ways we need to get to the point where we can have universal remote workers where anything that somebody does mediated by a computer, that doesn’t require physical interaction, an AGI will be able to do.” Carmack does not think this will be “unapproachably hard”: “That’s incredibly hubristic to say, but what I said a couple of years ago was that there’s a 50% chance that somewhere there will be signs of life of an AGI in 2030, and I’ve probably increased that slightly — maybe 55-60% now because I do think there’s a little sense of acceleration.” Video source: Lex Fridman (2022)](https://image.24vids.com/tw-2001635889982611555/amplify_video_thumb/2001612344477016064/img/T4Lkfg4IuehEJBTH.jpg)
