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

Sparks of Artificial General Intelligence: Early Experiments with GPT-4

arxiv.org

Post reply on HN
Print view View original

21–30 of 244 posts

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#21

Post by angusturner »

georgehill wrote: post #11
atleastoptimal wrote: post #6

Earlier quoted context omitted.

> It is hard to believe that our brain is just a bunch of attention layers and neural nets. Our brain isn't, but I'd wager the architectural complexity of a physical, neuronal brain is not optimized for all useful mental tasks, and has perhaps a fair amount of local maxima that are near vestigial in overall positive impact on cognition. Just because the human brain model of cognition is the only way nature has been a…

I agree that GI can have a different implementation compared to our human brain, but one thing is for sure: as of right now, the human brain can become more creative with a fraction of the data consumed by GPT-4. GPT-4 could be AGI, but it feels like cheating to achieve AGI by feeding the entire internet. If someone can build AGI with only the data that humans consume in their lifetime, then that, imho, is the real A…

I guess the challenge here is that the human mind is not a blank slate, and has been optimized first by billions of years of evolution.

If it takes all the data on the internet (or more) to bootstrap AGI, but that system is then capable of leveraging its knowledge to solve new out-of-distribution tasks, that seems like a fair test to me.

I agree with the article that we see "sparks" of this generality with GPT4.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#22

Post by maxdoop »

outlace wrote: post #14

ChatGPT and its relatives are very very impressive on first impressions, but I've been using ChatGPT-3 and now 4 heavily every day since they became available to individuals and once you start using them this much it becomes very clear how NOT intelligent they are. It really just seems like extremely impressive statistical inference after this much use and finding so many failure modes. But it is still impressive how…

Our of curiosity, what is GPT-4 getting wrong so often? It’s prettily wild to my own , admittedly easily impressed, mind.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#23

Post by steve_adams_86 »

outlace wrote: post #14

ChatGPT and its relatives are very very impressive on first impressions, but I've been using ChatGPT-3 and now 4 heavily every day since they became available to individuals and once you start using them this much it becomes very clear how NOT intelligent they are. It really just seems like extremely impressive statistical inference after this much use and finding so many failure modes. But it is still impressive how…

This has been my take as well. It's without a doubt at the point where it's useful and I'm very interested to experiment with it and see how much utility I can get out of it. At the same time, as much as I realize skill with prompting is essential, a lot of my time is spent programming automations around using those prompts. A lot of thought and care goes into it, and the failure rates of what I'm building are dramatically higher than anything I build without AI.

It's like I'm convincing AI to do a "smart" thing (which is really cool, don't get me wrong), but the AI truly has no idea what it's supposed to be doing. Sometimes even very direct, specific, concise prompting will yield totally abysmal responses.

I'm certainly not putting it down either though. It's unprecedented and it marks a genuinely important shift in the industry that this is possible at all. It just isn't all that smart when it comes to intents and actually knowing what's going on.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#24

Post by onos »

“ We note however that there is no single definition of AGI that is broadly accepted, and we discuss other definitions in the conclusion section.”

We know it can do a lot of cool stuff, but without a pinned down definition the headline here is useless.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#25

Post by ftxbro »

nycdatasci wrote: post #20

Can we agree that an "early (yet still incomplete) version of AGI" isn't AGI?

To me it's clear that they think it's AGI, and that they think AGI is such a loaded concept that it's not worth "announcing" it.

Beliefs are changing so fast right now. The term "AGI skeptic" will soon (if not already) mean "I don't trust AGIs in positions of authority or power" rather than "I don't think the technology is capable of matching our level of cognition."

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#26

Post by steve_adams_86 »

maxdoop wrote: post #22
outlace wrote: post #14

ChatGPT and its relatives are very very impressive on first impressions, but I've been using ChatGPT-3 and now 4 heavily every day since they became available to individuals and once you start using them this much it becomes very clear how NOT intelligent they are. It really just seems like extremely impressive statistical inference after this much use and finding so many failure modes. But it is still impressive how…

Our of curiosity, what is GPT-4 getting wrong so often? It’s prettily wild to my own , admittedly easily impressed, mind.

In my experience it has been failing to adhere to response formats. I can tell it to respond with answers in one of two specific formats, but the success rate is sometimes very low depending on the content of the inquiry it's supposed to respond to. Not sure why, it just seems to lose track of the response format.

I've been thinking maybe that's just the wrong way to approach it, yet it does work sometimes; it seems like it shouldn't be hard for the model to know how to respond when told explicitly how to do so.

It could be problems with my formatting, but the fact that it does work a lot of the time suggests it's something else.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#27

Post by steve_adams_86 »

adroniser wrote: post #7
georgehill wrote: post #2

> Given the breadth and depth of GPT-4’s capabilities, we believe that it could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system. I don't know why, but my brain refuses to accept GPT-4 as something close to AGI. Maybe I am wrong. It is hard to believe that our brain is just a bunch of attention layers and neural nets.

Do you think anything digital could ever become conscious?

I have a feeling it's like the saying "Any sufficiently advanced technology is indistinguishable from magic". At a certain point they could become practically identical things.

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#28

Post by TMWNN »

As a non-expert in the field I was hesitant at the time to disagree with the legions of experts who last year denounced Blake Lemoine and his claims. I know enough to know, though, of the AI effect https://en.wikipedia.org/wiki/AI_effect>, a longstanding tradition/bad habit of advances being dismissed by those in the field itself as "not real AI". Anyone, expert or not, in 1950, 1960, or even 1970 who was told that before the turn of the century a computer would defeat the world chess champion would conclude that said feat must have come as part of a breakthrough in AGI. Same if told that by 2015 many people would have in their homes, and carry around in their pockets, devices that can respond to spoken queries on a variety of topics.

To put another way, I was hesitant to be as self-assuredly certain about how to define consciousness, intelligence, and sentience—and what it takes for them to emerge—as the experts who denounced Lemoine. The recent GPT breakthroughs have made me more so.

I found this recent Sabine Hossenfelder video interesting. https://www.youtube.com/watch?v=cP5zGh2fui0>

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#29

Post by throwaway4837 »

outlace wrote: post #14

ChatGPT and its relatives are very very impressive on first impressions, but I've been using ChatGPT-3 and now 4 heavily every day since they became available to individuals and once you start using them this much it becomes very clear how NOT intelligent they are. It really just seems like extremely impressive statistical inference after this much use and finding so many failure modes. But it is still impressive how…

Simple thought experiment: if I was a high school teacher and graded hundreds of exams, papers, and homework every month for years, would I develop the same opinion of my C-level students? The ones who have somewhat correct work, often inconsistent, clear that they're not thinking about the material, etc.

Humans all start out like this, but as they grow they develop different masteries. Many of those masteries are easy to learn and shared across all humans, like how to walk or tie your shoes. Many are domain-specific, like how to perform heart surgery or solve complex math problems.

I don't think AGI is a well-defined term based on the way it's thrown around, but sure I don't think it's "AGI". I do know it's going to be very useful technology, especially multimodal GPT-4 and beyond. I think our interface with AI is going to be multimodal.

Suddenly Google Glass could be an interesting product again. Glasses would stream audio/video to your phone, you'd be able to ask questions about things you're seeing, and it would display the answers on the screen or dictate them back into your earbuds. I like the idea of having a travel companion, another brain at my fingertips. Sort of like what the mobile phone did for us, but a lot less work for the user. I'm getting impatient waiting for this GPT-4 API access!

Re: Sparks of Artificial General Intelligence: Early Experiments with GPT-4

#30

Post by chrischen »

outlace wrote: post #14

ChatGPT and its relatives are very very impressive on first impressions, but I've been using ChatGPT-3 and now 4 heavily every day since they became available to individuals and once you start using them this much it becomes very clear how NOT intelligent they are. It really just seems like extremely impressive statistical inference after this much use and finding so many failure modes. But it is still impressive how…

It’s almost as if someone who can piece together words well enough they can make people think they are smarter than they actually are.
Post reply on HN