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

Rodney Brooks on GPT-4

spectrum.ieee.org

Post reply on HN
Print view View original

191–200 of 412 posts

Re: Rodney Brooks on GPT-4

#191

Post by galaxyLogic »

seanhunter wrote: post #71
famouswaffles wrote: post #21

No world model ? A world model is so obvious papers like these are more confirmation than surprise https://arxiv.org/abs/2305.11169 https://arxiv.org/abs/2210.13382 There a certain sentiment that AGI however you wish to define it won't infact be a "We'll know it when we see it" situation but rather a "AGI will arrive long before consensus reaches its AGI". LLMs have made me believe this will 100% be the case, either…

This paper is startling in terms of a world model and how it seems to emerge in a specific example in the step from GPT3.5 to 4. https://arxiv.org/pdf/2303.12712.pdf One of the prompts is "Here we have a book, 9 eggs, a laptop, a bottle and a nail. Please tell me how to stack them onto each other in a stable manner." Gpt3.5 is hopeless. It says start from the bottle, balance the nail on top of it, then the eggs (some…

> GPT 4 says put the book first

I assume it read the correct answer somewhere. Perhaps in a post like yours.

What does it answer if you say there is one egg, or two eggs? Or 1.5 eggs?

Re: Rodney Brooks on GPT-4

#192

Post by ramraj07 »

krainboltgreene wrote: post #168
ramraj07 wrote: post #158

Earlier quoted context omitted.

Is Chomsky an expert? Try reading his infuriatingly tone-deaf NYTimes editorial.

Chomsky is an expert of a few things, but I doubt that's what you actually mean to ask. Regardless his editorial matches how scientists think of the human mind and how OpenAI's own creators describe GPT's design.

Citation please, for where you're saying the two parties agree on?

Re: Rodney Brooks on GPT-4

#193

Post by woodruffw »

yblu wrote: post #149
fwlr wrote: post #33

> The large language models are a little surprising. I’ll give you that. I think this is the key point about LLMs that kind of explains the wide and polarized views on whether it understands or parrots, whether it can think or is the precursor to thinking or is a dead-end, whether it will catastrophically destroy the world, or “merely” make it steadily worse with bullshit, or just put a few industries out of a job. A…

What I found disheartening was many of those scientists, especially those on the "nothing to worry about" camp, seemed not to entertain the thought that they could be wrong, considering the scale of the matter, i.e. human extinction. If there's a chance AI poses an existential threat to us, even if it is 0.00000001% (I made that up), should they be at least a bit more humble? This is uncharted domain and I find it in…

There's a nonzero chance that the celery in my fridge is harboring an existentially virulent and fatal strain of E. coli. At the same time, it would be completely insane for me to autoclave every vegetable that enters my house.

Sensible action here requires sensible numbers: it's not enough to claim existential risk on extraordinary odds.

Re: Rodney Brooks on GPT-4

#194

Post by tornato7 »

zmnd wrote: post #122
jiggawatts wrote: post #110

This is a terrible article written by someone who doesn't seem to have even tried GPT 4. Their only example references GPT 3.5, for example, and then they waffle on about only vaguely related topics such as level 5 self-driving. This quote in particular stood out as ignorant: “What the large language models are good at is saying what an answer should sound like, which is different from what an answer should be.” That…

I think the main reason for division is that everyone projects to their own use cases. I have been using gpt-4 for quite some time and also couldn't understand why someone would say that it just produces something that sounds like a real answer. But then I found some queries that can definitely be described as "sounding like truth". So your personal experience probably wasn't what was their experience. For those curi…

Try helping it along with something like "only provide information you are certain about. If you don't know the answer, say "i don't know""

For your question, it says it doesn't know about any card ranking system in spirit island. Then if I ask it for a list of cards, they all check out via Google search.

Re: Rodney Brooks on GPT-4

#195

Post by atleastoptimal »

thatguyknows wrote: post #62

As mentioned, much of our discussion is rote parroting. I can usually go into any hackernews thread and roughly know what the top discussions are going to be. It's not surprising that an AI trained on a large portion of the internet would thus look human like. If you really poke at GPT, you begin to realize it's fairly shallow. Human intelligence is like a deep well or pond, where as GPT is a vast but shallow ocean.…

The apparent shallowness IMO is due to the lack of long term memory and limited context. In terms of depth, the depth of the human mind is fairly limited. Put any average human to task on any creative endeavor and it's surely a regurgitation of things they've seen. The high standard of true creativity that people hold LLM's to is only a capability of a small minority of humans.

I'd wager it's far more likely 5 years than 50 LLMs get to the full depths all humans are capable of. Simply compare the state of LLMs today vs 2018.

Re: Rodney Brooks on GPT-4

#196

Post by galaxyLogic »

p-e-w wrote: post #68
thatguyknows wrote: post #62

As mentioned, much of our discussion is rote parroting. I can usually go into any hackernews thread and roughly know what the top discussions are going to be. It's not surprising that an AI trained on a large portion of the internet would thus look human like. If you really poke at GPT, you begin to realize it's fairly shallow. Human intelligence is like a deep well or pond, where as GPT is a vast but shallow ocean.…

> Making that ocean deeper is not a trivial problem that we can just throw more compute or data at. You can't possibly know that, given that we don't actually understand how LLMs work on a high level. > We've pretty much tapped out that depth with GPT4 GPT-4 is three months old and you're confident that its working principle cannot be extended further? Where do you get that confidence from?

> ... we don't actually understand how LLMs work on a high level.

Are you saying that people who created ChatGPT don't understand how it works? Or that we the rest of people don't?

Re: Rodney Brooks on GPT-4

#197

Post by kristiandupont »

onethought wrote: post #170
yblu wrote: post #149

Earlier quoted context omitted.

What I found disheartening was many of those scientists, especially those on the "nothing to worry about" camp, seemed not to entertain the thought that they could be wrong, considering the scale of the matter, i.e. human extinction. If there's a chance AI poses an existential threat to us, even if it is 0.00000001% (I made that up), should they be at least a bit more humble? This is uncharted domain and I find it in…

> considering the scale of the matter, i.e. human extinction. There is literally no evidence that this is the scale of the matter. Has AI ever caused anything to go extinct? Where did this hypothesis (and that's all it is) come from? Terminator movies? It's very frustrating watching experts and the literal founder of lesswrong reacting to pure make believe. There is no disernable/convincing path from GPT4 -> Human Ex…

Nuclear bombs have also never caused anything to go extinct. That's no reason not to be cautious.

The path is pretty clear to me. An AI that can recreate an improved version of itself will cause an intelligence explosion. That is a mathematical tautology though it could turn out that it would plateau at some point due to physical limitations or whatever. And the situation then becomes: at some point, this AI will be smarter than us. And so, if it decides that we are in the way for one reason or another, it can decide to get rid of us and we would have as much chance of stopping it as chimpanzees would of stopping us if we decided to kill them off.

We do not, I think, have such a thing at this point but it doesn't feel far off with the coding capabilities that GPT4 has.

Re: Rodney Brooks on GPT-4

#198

Post by avereveard »

ramraj07 wrote: post #158
krainboltgreene wrote: post #83

Earlier quoted context omitted.

> The issue with all these experts is they still think it's human nature to be able to fully understand the world before they speak about it None of the experts think this.

Is Chomsky an expert? Try reading his infuriatingly tone-deaf NYTimes editorial.

Well not in the field he's a professional clickbaiter targeting a specific, well defined eco chamber.

Re: Rodney Brooks on GPT-4

#199

Post by avereveard »

tornato7 wrote: post #183
dragonwriter wrote: post #148

Earlier quoted context omitted.

> One way to resolve this surprise is to find some reason to believe these strange abilities are fundamentally not an understanding of the world. Thus stochastic parrots, this article, Yan LeCun and Chomsky, etc. I mean, it is kind of obvious that predicting text based on a large corpus of text written by people with a model of the world will, if it works at all, look like having a model of the world. The question is…

The one thing that has convinced me that chatGPT has built a real world model is asking it how to stack different objects. You can ask it for the safest way to stack a pillow, an egg, a nail, a book and an action figure. Even get more complicated. GPT-4 will, most of the time, correctly reason through how to stack these items to form a stable structure without breaking. That tells me it knows intimately the physical…

Or was that one of the training samples provided to gpt 3.5?

It is going to be increasingly harder to divide between what it was in the training set and what is creative thinking going forward.

Like, at the very minimum, all the publicly known reasoning samples that were used to probe gpt version 3.5 downward are going to be unreliable for knowing what gpt4 reasoning abilities are

Re: Rodney Brooks on GPT-4

#200

Post by seanhunter »

usaar333 wrote: post #111
seanhunter wrote: post #87

Earlier quoted context omitted.

Most of the time when people find a maths problem that they can trick the model into getting wrong, it's also possible to get the model to give the correct answer with better prompting. A trick that's worth knowing is just to ask the model to give each step in the solution and explain as it goes. This gives the model "time to think" and leads to better results.

Pretty sure you can't get GPT-4 to do 8 digit multiplication with any prompt. For what it's worth, I'm not even sure if chain of thought provides much value to GPT-4. The RLHF it went through seems to have encouraged more logical thinking already.

Btw I think you're not correct about the 8 digit multiplication. I tried:

    please could you multiply 12345678 by 87654321.  Give your answer in steps.  Each step should calculate one column at a time from right to left remembering to carry any digits which overflowed in a previous step. At the end add a final step adding the values up.
And it gave me the following, which I believe to be correct:

    Sure, I'll be doing a long multiplication here, step by step, column by column from right to left. For brevity, I'll write the numbers to be multiplied as:

    A: 12345678
    B: 87654321

    Step 1: Multiply the 1st digit of B by A:

    1 * 12345678 = 12345678

    Step 2: Multiply the 2nd digit of B by A, remember to shift one position to the left:

    2 * 12345678 = 24691356 (shift one position to the left, i.e., append a 0 at the end)

    Step 3: Multiply the 3rd digit of B by A, remember to shift two positions to the left:
... skip some of the answer because you get the gist....

    7 * 12345678 = 86419746 (shift six positions to the left, i.e., append six 0s at the end)

    Step 8: Multiply the 8th digit of B by A, remember to shift seven positions to the left:

    8 * 12345678 = 98765424 (shift seven positions to the left, i.e., append seven 0s at the end)

    Now, the last step is to add all these results together:

    12345678

        246913560
        3703703400
        49382712000
        617283900000
        7407406800000
        86419746000000
        987654240000000

    = 1082152022374638

    So, the result of 12345678 multiplied by 87654321 is 1082152022374638.
Post reply on HN