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

361–370 of 412 posts

Re: Rodney Brooks on GPT-4

#361

Post by YeGoblynQueenne »

jiggawatts wrote: post #269
omnicognate wrote: post #241

Earlier quoted context omitted.

I've been paying for GPT-4 since it came out and have used it extensively. It's clearly an iteration on the same thing and behaves in qualitatively the same way. The differences are just differences of degree. It's not hard to get a feel for the "edges" of an LLM. You just need to come up with a sequence of related tasks of increasing complexity. A good one is to give it a simple program and ask what it outputs. Then…

Most developers -- let alone humans -- I've met can't run trivial programs in their head successfully, let alone complex ones. I've thrown crazy complicated problems at GPT 4 and had mixed results, but then again, I get mixed results from people too. I've had it explain a multi-page SQL query I couldn't understand myself. I asked it to write doc-comments for spaghetti code that I wrote for a programming competition,…

My take-away is that your interaction with the OP has not changed your opinion about "everyone", expressed above:

>> What I and many others have noticed about the "Are LLMs really smart?" debate is that everyone on the "Nay" side is using 3.5 and everyone on the "Yay" side is using 4.0.

Sometimes there really is no point in trying to make curious conversation. Curiosity has left the building.

Re: Rodney Brooks on GPT-4

#362

Post by furyofantares »

coldtea wrote: post #96
furyofantares wrote: post #75

Earlier quoted context omitted.

Alright, I give. What the heck is "sapience"?

You know, the capacity of though/intelligence/understanding the world? As in "homo sapiens" (same latin root)

Sounds by a wide margin even less well defined than AGI or sentience

Re: Rodney Brooks on GPT-4

#363

Post by thomastjeffery »

zainhoda wrote: post #172
thomastjeffery wrote: post #107

Earlier quoted context omitted.

> My current (tentative) resolution of the surprise is that language encoded way more information about reality than we thought it did. I think you are close to the mark, but you have been subtly mislead: language is not the data we are working with. We are working with text . Once you fix that particular failure of word choice, everything else becomes much more clear: text contains much more information than languag…

There’s an exercise that some people do when learning programming, which is to write down the steps to make a sandwich. Then the teacher follows the exact instructions to make a sandwich and most people don’t put enough detail for a computer to follow (I.e. open the fridge etc) and the teacher will run around bumping into things. That used to be a teaching exercise to show people the amount of precision required when…

> Now with LLMs, I think one of the great leaps is the idea that it’s no longer necessary to be “pedantic” when giving computers instructions

Yes, but they also can't. They can't be pedantic or follow explicit instructions. That's the other side of the coin that isn't being presented.

They can present the right elements of the story in the the right places, but they can't perform it.

Re: Rodney Brooks on GPT-4

#364

Post by thomastjeffery »

tornato7 wrote: post #178
thomastjeffery wrote: post #107

Earlier quoted context omitted.

> My current (tentative) resolution of the surprise is that language encoded way more information about reality than we thought it did. I think you are close to the mark, but you have been subtly mislead: language is not the data we are working with. We are working with text . Once you fix that particular failure of word choice, everything else becomes much more clear: text contains much more information than languag…

One experiment that I would love to see is an LLM-like model for audio. Feed it hours and hours of lectures, sound effects, animal calls, music etc. You would be able to talk to it and it would ingest the raw waveform then produce audio as a response. Would it learn the fundamentals of music theory? Would it learn to produce "the sound of a bowling ball hitting a dozen windchimes?" Would it learn to talk in English a…

> Would it learn the fundamentals of music theory?

No, but you might convince yourself it did.

It would map the patterns that exist in its training set. It would then follow those patterns. The result would look like a human understanding music theory, but it would not be that.

It would be stumbling around exactly the domain we gave it: impressive because that domain is not noise, it's good data. It still wouldn't be able to find its way around, only stumble.

Re: Rodney Brooks on GPT-4

#365

Post by pixl97 »

omnicognate wrote: post #301
jiggawatts wrote: post #269

Earlier quoted context omitted.

Most developers -- let alone humans -- I've met can't run trivial programs in their head successfully, let alone complex ones. I've thrown crazy complicated problems at GPT 4 and had mixed results, but then again, I get mixed results from people too. I've had it explain a multi-page SQL query I couldn't understand myself. I asked it to write doc-comments for spaghetti code that I wrote for a programming competition,…

"In a distinctly non-humanlike way ". You can learn a lot about how a system works from how it fails and in this case it fails in a way consistent with the token-prediction approach we know it is using rather than the model-forming approach some are claiming has "emerged" from that. It doesn't show the performance on a marginally more complex example that you would expect from a human with the same performance on the…

>"In a distinctly non-humanlike way".

Over the long term this is going to be a primary alignment problem of AI as it becomes more capable.

What is my reasoning behind that?

Because humans suck, or at least our constraints that we're presented with do. All your input systems to your brain are constantly behind 'now' and the vast majority of data you could input is getting dropped on the ground. For example if I'm making a robotic visual input system, it makes nearly zero sense for it to behave like human vision. Your 20/20 visual acuity area is tiny and only by moving your eyes around rapidly and then by your brain lying to you, do we have a high resolution view on the world.

And that is just an example of one of those weird human behaviors we know about. It's likely we'll find more of these shortcuts over time because AI won't take them.

Re: Rodney Brooks on GPT-4

#366

Post by goatlover »

JoshTko wrote: post #38
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…

The more I think about it the more I'm convinced I am basically just predicting/saying my next word whenever I speak.

Does that reasoning work when applied to what you just wrote? If you’re just predicting the next word, then what makes you confident in the conclusion that you’re just predicting your next word?

Re: Rodney Brooks on GPT-4

#367

Post by satvikpendem »

zainhoda wrote: post #172
thomastjeffery wrote: post #107

Earlier quoted context omitted.

> My current (tentative) resolution of the surprise is that language encoded way more information about reality than we thought it did. I think you are close to the mark, but you have been subtly mislead: language is not the data we are working with. We are working with text . Once you fix that particular failure of word choice, everything else becomes much more clear: text contains much more information than languag…

There’s an exercise that some people do when learning programming, which is to write down the steps to make a sandwich. Then the teacher follows the exact instructions to make a sandwich and most people don’t put enough detail for a computer to follow (I.e. open the fridge etc) and the teacher will run around bumping into things. That used to be a teaching exercise to show people the amount of precision required when…

It depends on the task. For certain programs you absolutely need to be pedantic in describing what needs to happen. There is a reason we don't program in natural language and that won't change with LLMs.

Re: Rodney Brooks on GPT-4

#368

Post by satvikpendem »

thomastjeffery wrote: post #364
tornato7 wrote: post #178

Earlier quoted context omitted.

One experiment that I would love to see is an LLM-like model for audio. Feed it hours and hours of lectures, sound effects, animal calls, music etc. You would be able to talk to it and it would ingest the raw waveform then produce audio as a response. Would it learn the fundamentals of music theory? Would it learn to produce "the sound of a bowling ball hitting a dozen windchimes?" Would it learn to talk in English a…

> Would it learn the fundamentals of music theory? No, but you might convince yourself it did. It would map the patterns that exist in its training set. It would then follow those patterns. The result would look like a human understanding music theory, but it would not be that . It would be stumbling around exactly the domain we gave it: impressive because that domain is not noise, it's good data. It still wouldn't b…

> The result would look like a human understanding music theory, but it would not be that.

The question then becomes, what is understand? Is what a human does any different than what this LLM is doing?

Re: Rodney Brooks on GPT-4

#369

Post by famouswaffles »

galaxyLogic wrote: post #196
p-e-w wrote: post #68

Earlier quoted context omitted.

> 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?

Training a model doesn't mean you understand what the neurons actually do to influence output. Nobody knows that. That's where the black box analogies come in. We know what goes in the box and what comes out. We don't know what the box is doing to the data

Re: Rodney Brooks on GPT-4

#370

Post by krainboltgreene »

leobg wrote: post #304
biorach wrote: post #302

Earlier quoted context omitted.

> Now ChatGPT is pretty much Skinner’s work come to life. I'm pretty sure that is based on a misunderstanding of Skinner, ChatGPT, or both

Or very lossy compression of what I mean. I have studied both quite extensively. Not 10,000 hours each, but hundreds for sure.

The difference between hundreds and 10k hours is roughly 10k hours.
Post reply on HN