Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

dwarkesh.com

Post reply on HN
Print view View original

321–330 of 394 posts

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#321

Post by imtringued »

timewizard wrote: post #171
yibg wrote: post #162

Might as well be 10 - 1000 years. Reality is no one knows how long it'll take to get to AGI, because: 1) No one knows what exactly makes humans "intelligent" and therefore 2) No one knows what it would take to achieve AGI Go back through history and AI / AGI has been a couple of decades away for several decades now.

That we don't have a single unified explanation doesn't mean that we don't have very good hints, or that we don't have very good understandings of specific components. Aside from that the measure really, to me, has to be power efficiency. If you're boiling oceans to make all this work then you've not achieved anything worth having. From my calculations the human brain runs on about 400 calories a day. That's an absur…

Energy efficiency is not really a good target since you can brute force it by distilling classical ANNs to spiking neural networks.

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#322

Post by dmwilcox »

krisoft wrote: post #265
dmwilcox wrote: post #251

Earlier quoted context omitted.

I tried to keep my long post short so I cut things. I gestured at it -- there is nothing in a computer we didn't put there. Take the same model weights give it the same inputs, get the same outputs. Same with the pseudo-random number generator. And the "same inputs" is especially limited versus what humans are used to. What's the machine code of an AGI gonna look like? It makes one illegal instruction and crashes? If…

> What's the machine code of an AGI gonna look like? Right now the guess is that it will be mostly a bunch of multiplications and additions. > It makes one illegal instruction and crashes? And our hearth quivers just slightly the wrong way and we die. Or a tiny blood cloth plugs a vessel in our brain and we die. Do you feel that our fragility is a good reason why meat cannot be intelligent? > I jest but really think…

Mmmm well my meatsuit can't easily make my own heart quiver the wrong way and kill me. Computers can treat data as code and code as data all pretty easily. It's core to several languages (like lisp). As such making illegal instructions or violating the straightjacket of a system such an "intelligence" would operate in is likely. If you could make an intelligent process, what would it think of an operating system kernel (the thing you have to ask for everything, io memory, etc)? Does the "intelligent" process fear for itself when it's going to get descheduled? What is the bitpattern for fear? Can you imagine an intelligent process in such a place, as static representation of data in ram? To get write something down you call out to a library and maybe the CPU switches out to a brk system call to map more virtual memory? It all sounds frankly ridiculous. I think AGI proponents fundamentally misunderstand how a computer works and are engaging in magical thinking and taking the market for a ride.

I think it's less about the randomness and more about that all the functionality of a computer is defined up front, in software, in training, in hardware. Sure you can add randomness and pick between two paths randomly but a computer couldn't spontaneously pick to go down a path that wasn't defined for it.

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#323

Post by imtringued »

ksec wrote: post #24

Is AGI even important? I believe the next 10 to 15 years will be Assisted Intelligence. There are things that current LLM are so poor I dont believe a 100x increase in pref / watt is going to make much difference. But it is going to be good enough there wont be an AI Winter. Since current AI has already reached escape velocity and actually increase productivity in many areas. The most intriguing part is if Humanoid f…

I am thinking of designing machines to be used in a flexible manufacturing system and none of them will be humanoid robots. Humanoid robots suck for manufacturing. They're walking on a flat floor so what the heck do they need legs for? To fall over?

The entire point of the original assembly line was to keep humans standing in the same spot instead of wasting time walking.

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#324

Post by imtringued »

AstroBen wrote: post #192
Retric wrote: post #170

Earlier quoted context omitted.

The general part of general intelligence. If they don’t think in those terms there’s an inherent limitation. Now, something that’s arbitrarily close to AGI but doesn’t care about endlessly working on drudgery etc seems possible, but also a more difficult problem you’d need to be able to build AGI to create.

Artificial general intelligence (AGI) refers to the hypothetical intelligence of a machine that possesses the ability to understand or learn any intellectual task that a human being can. Generalization ability and Common Sense Knowledge [1] If we go by this definition then there's no caring, or a noticing of drudgery? It's simply defined by its ability to generalize solving problems across domains. The narrow AI that…

Exactly. It's called artificial general intelligence, not human general intelligence.

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#325

Post by imtringued »

gloosx wrote: post #315
ggreer wrote: post #237

Earlier quoted context omitted.

Is there any specific mental task that an average human is capable of that you believe computers will not be able to do? Also does this also mean that you believe that brain emulations (uploads) are not possible, even given an arbitrary amount of compute power?

1. Computers cannot self-rewire like neurons, which means human can pretty much adapt for doing any specific mental task (an "unknown", new task) without explicit retraining , which current computers need to learn something new 2. Computers can't do continuous and unsupervised learning, which means computers require structured input, labeled data, and predefined objectives to learn anything. Humans learn passively al…

Minor nitpicks. I think your points are pretty good.

1. Self-rewiring is just a matter of hardware design. Neuromorphic hardware is a thing.

2. LLM foundation models are actually unsupervised in a way, since they simply take any arbitrary text and try to complete it. It's the instruction fine-tuning that is supervised. (Q/A pairs)

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#327

Post by xbmcuser »

Most people talking about Ai and economic growth have vested interests in talking about how it will increase economic growth but don't talk about that under current economic system that the world has would mean most if not all of the growth will go to > 0.0001% of the population.

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#328

Post by ceoofballin »

TURKS MENTIONED RAAAAAAAAAAAAAAAHHHHH!!1!1!! TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE TURKIYE

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#329

Post by harshitaneja »

olyjohn wrote: post #206
kadushka wrote: post #121

Earlier quoted context omitted.

What does intelligence have to do with having desires or goals? An amoeba can do stuff on its own but it's not intelligent. I can imagine a god-like intelligence that is a billion times smarter and more capable than any human in every way, and it could just sit idle forever without any motivation to do anything.

Does the amoeba does make choices?

Do you make that choice? Did I?

Re: AGI Is Still 30 Years Away – Ege Erdil and Tamay Besiroglu

#330

Post by Retric »

AstroBen wrote: post #305
Retric wrote: post #296

Earlier quoted context omitted.

> Evaluating emotional context would be similar to a chess engine calculating its next move. There's nothing there that implies a conscience, sentience, morals, feelings, suffering or anything 'human'. It's just a necessary intermediate function to achieve its goal If it’s limited to achieving goals it’s not AGI. Real time personal goal setting based on human equivalent emotions is an “intellectual task.” One of many…

I mean we just fundamentally have different definitions of AGI. Mine's based on outcomes and what it can do, so purely goal based. Not the processes that mimic humans or animals I think this is the most likely first step of what would happen seeing as we're pushing for it to be created to solve real world problems

I’m not sure how you can argue something is a general intelligence if it can’t do those kinds of things? Comes out of the factory with a command: “Operate this android for a lifetime pretending to be human.”

Seems like arguing something is a self driving car if it needs a backup human driver for safety. It’s simply not what people who initially came up with the term meant and not what a plain language understanding of the term would suggest.

Post reply on HN