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

131–140 of 244 posts

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

#131

Post by hallqv »

number6 wrote: post #104
hallqv wrote: post #90

Been pair-coding with gpt4 for the last week, it's definitely AGI..

How do you pair program with ChatGPT?

Depends on the task but some combination of asking it for skeleton code for new tasks and sending it my written code or error messages and asking for corrections or potential solutions.

It's very effective, if you are atleast semi-new to technology you are using it will explain and teach you things you didn't know before, and if you know the tech by heart it saves you from having to type it out.

For example, yesterday I had to make a custom container with some pretty involved dependiencies that also had to be be runnable on AWS Lambda (which I haven't used much before), me and gpt4 went back and forth with Dockerfile code and error messages for a few hours and then it ran like charm. Would probably have taken me 1-2 days of regular coding and googling otherwise.

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

#132

Post by jhugo »

ccozan wrote: post #128
hislaziness wrote: post #79

For whatever reason we seem to have set a very high expectation from AI as compared to NI (Natural Intelligence). I remember reading "There is a considerable overlap between the intelligence of the smartest bears and the dumbest tourists."

An AI that is just a chat or some LLM is not going to be too relevant for human life ( thanks, I can google stuff or ask a friend; also writing poems is just fun, but not of any usefulness ). But where are my damn robots that I can assign task and do them reliably ( clean the garden, go get this list of groceries - or , just look in the damn fridge and go buy what is missing , and so on )? Then AI is useful.

> But where are my damn robots that I can assign task and do them reliably ( clean the garden, go get this list of groceries - or , just look in the damn fridge and go buy what is missing , and so on )?

The amusing/terrifying thing is that AI is probably going to replace a bunch of people's jobs before it can take away any of that menial labour from them.

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

#133

Post by cowl »

IanCal wrote: post #117
cowl wrote: post #71

Earlier quoted context omitted.

So you predict the top comments For a claim would be: 1. Dismisal 2. Trivialism 3. Non Well Formed Claim 4. I accept the claim 5. Disprove by counter example Are you sure you have not forgotten any tactic of debate to include in you prediction? I predict that you Prediction will result probably in these actions: 1. upvoted 2. downvoted

6. Discussion around the idea that adds something of value.

Every one of them adds value. Question the premises (wellformed question) Question the motives/importance (dismisal) Disprove by counter example

The only one that can be questioned if it adds something of value is the acceptance and surrender one (I welcome the overlords) although in some forms of thought, advising t0 not fight with the inevitable can be viewed as constructive.

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

#134

Post by hallqv »

atleastoptimal wrote: post #102
hallqv wrote: post #90

Been pair-coding with gpt4 for the last week, it's definitely AGI..

No it's a Chinese room but instead of Chinese it's stack overflow snippets

So what? If it writes novels like an AGI, codes like an AGI and explains complex topics like an AGI, then it's probably an AGI...

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

#136

Post by ilitirit »

GPT AI systems remind me of Chinese Room thought experiment:

https://en.wikipedia.org/wiki/Chinese_room

This is also similar to the Duck Test:

https://en.wikipedia.org/wiki/Duck_test

Depending on the context, there are generally two takes: "It is (or is not) a duck", and "It doesn't (or does) matter whether or not it's a duck". These aren't mutually exclusive.

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

#137

Post by ofrzeta »

js8 wrote: post #123
blueorange8 wrote: post #106

Earlier quoted context omitted.

That Goat wolf cabbage problem gpt-4 can solve already

Yes, but does it because it read somewhere? Can it adapt the existing solution to a new variation? Can it solve a similar problem with different things? This is what humans do all the time.

It can solve the variation "You have a rabbit, a wolf, a haypile and you want to cross a chasm". What kinds of variations do you have in mind?

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

#138

Post by templeosenjoyer »

Unless they somehow cured GPT-3's schizophrenia and this model is a significant upgrade I'm not buying it - no matter how good it is at proving trivial mathematics theorems in the style of Eliot or whoever. Too often I have dealt with "The answer to your question is X. Oh, sorry, you are right, the answer is actually Y. Oh, it is good of you to ask for a proof, sure I can prove the answer is Y, I used this (hallucinated) method described in this (hallucinated) paper. Oh, sorry, you are right, I cannot find any evidence that the method and paper I mentioned earlier actually exist, oops!".

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

#139

Post by js8 »

ofrzeta wrote: post #137
js8 wrote: post #123

Earlier quoted context omitted.

Yes, but does it because it read somewhere? Can it adapt the existing solution to a new variation? Can it solve a similar problem with different things? This is what humans do all the time.

It can solve the variation "You have a rabbit, a wolf, a haypile and you want to cross a chasm". What kinds of variations do you have in mind?

If I add a small condition that makes the solution impossible, will it recognize that? Will it recognize for your example that it's a variation? Will it still be able to solve it when it is just a subtask of a bigger input?

If I ask it a leading question that intentionally relies on a wrong solution, will it recognize that?

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

#140

Post by SanderNL »

IanCal wrote: post #110

I'm increasingly convinced you can build an agi system with gpt4. People are trying to get it to solve everything up front but I've had GPT3 do much better by taking it through a problem asking it questions. Then I realised it was good at asking those questions too so just hooked it up to talk to itself with different roles. Gpt4 seems much better overall and is very good at using tools if you just tell it how and wh…

I agree. There is something special about layering these guys. To me this is like we are looking at a static combustion engine without the vehicle. “How is this useful?”

It’s that I’m not sure what the best approach is here. Waiting for other smarter folks to put the pieces together.

Post reply on HN