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

131–140 of 412 posts

Re: Rodney Brooks on GPT-4

#131

Post by astrea »

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.

But you and your brain are much more than the language-forming aspect.

Re: Rodney Brooks on GPT-4

#132

Post by gmerc »

thatguyknows wrote: post #72
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?

Sam Altman said it himself. He seems like a reasonable source. If you're familiar with other fields of AI, adding more and more layers to ResNet was the hotness for awhile, but the trick stopped working after awhile.

Sure Sam Altman, the lying CEO of a tech company (they all do) should be listened to on this matter but not on the part where he thinks AGI within reach using his approach. Selective hearing.

Re: Rodney Brooks on GPT-4

#133

Post by nirav72 »

usaar333 wrote: post #109
mrmincent wrote: post #13

I get why AI people are at pains to say that GPT-* isn’t AI, and that agi is still a long, difficult way off, I do understand that the distinction is important. But ChatGPT has become such a useful tool to help explain a concept or to filter thoughts through I don’t really care if it’s proper AI or just playing pretend. Google search gives me pretty useless results these days, forums are slow and inconsistent to resp…

AGI is just a poorly defined, moving target. Absent the goals constantly shifting, GPT-3 can be viewed as one, GPT-4 even more so. You can ask it questions about almost anything (at a broad level) and get an answer. That's what makes it general and "intelligent"

>makes it general and "intelligent"

Isn't intelligent a matter of perspective? Most people that are critical of GPT-4 wonder if it ever produces anything novel. Since its been trained on existing text created by humans. So it's replicating those patterns in its output. But yes, it has its general purpose use as a tool. But it has its limits. Just the other day, there was an article posted on HA about how LLM's can't handle negation and tend to fall apart.

Here is the article. https://www.quantamagazine.org/ai-like-chatgpt-are-no-good-a...

Re: Rodney Brooks on GPT-4

#134

Post by gmerc »

tunesmith wrote: post #11

It's funny how it's possible to simultaneously overestimate and underestimate GPT4 at the same time, vastly. I think that we just don't fully understand everything it gives us yet. The complaints of "well it explained this wrong" are over-emphasized. The same thing happens with google and with any sort of research. Besides, if you're actually being productive with GPT4, you're going to be asking it stuff that relates…

Every time there is a thread in the topic humans are hallucinating and making shit up at epic scale. So really, it’s either a terminal flaw and we should never listen to another human again or we should start getting some humility

Re: Rodney Brooks on GPT-4

#135

Post by dragonwriter »

jiggawatts wrote: post #123
sheeshkebab wrote: post #5

Earlier quoted context omitted.

Anytime I ask these things something (bard, gpt etc), 33% of the answer is genius, 33% misleading garbage, 33% filler stuff that’s neither here or there The problem is distinguishing between these parts requires me to be be an expert in the area I’m inquiring about - and then why the heck do I need to ask some idiot bot for answers to questions that I already know an answer to? I don’t know who finds these things use…

Bard is a brain-damaged-but-literate idiot compared to GPT 3, which is still dumber than the typical human. Try GPT 4 for a week. I've found it to be more like 50% immediately useful, 25% very impressive, and 25% where it's not wrong but I have to poke it a few times with different prompts to coax out the specific answer I'm looking for. That's better than most humans that I collaborate with at work. Literally half o…

> I’ve had coworkers fail to understand words like “orthogonal” or “vanilla” in a sentence. Vanilla!

Presumably, you are referring to the idiomatic use of vanilla, which is probably a less-universal idiom than you think it is (it is of fairly recent origin in wide use, derives from a specific American cultural loading of the literal vanilla flavor) and which, even when the general idiom is understood, can rely on a deeply shared understanding of what is the basic default in the referenced context to actually be understand as to its contextual meaning.

Re: Rodney Brooks on GPT-4

#136

Post by ChatGTP »

jiggawatts wrote: post #123
sheeshkebab wrote: post #5

Earlier quoted context omitted.

Anytime I ask these things something (bard, gpt etc), 33% of the answer is genius, 33% misleading garbage, 33% filler stuff that’s neither here or there The problem is distinguishing between these parts requires me to be be an expert in the area I’m inquiring about - and then why the heck do I need to ask some idiot bot for answers to questions that I already know an answer to? I don’t know who finds these things use…

Bard is a brain-damaged-but-literate idiot compared to GPT 3, which is still dumber than the typical human. Try GPT 4 for a week. I've found it to be more like 50% immediately useful, 25% very impressive, and 25% where it's not wrong but I have to poke it a few times with different prompts to coax out the specific answer I'm looking for. That's better than most humans that I collaborate with at work. Literally half o…

Not to be rude, but are you paid by OpenAI to say this?

The amount of comments here telling people to “upgrade to ChatGPT 4” is absolutely unprecedented.

I know it might be good, but people will find value in it and upgrade if they see the need to do so?

Re: Rodney Brooks on GPT-4

#137

Post by gmerc »

ChatGTP wrote: post #129
PsylentKnight wrote: post #45

Earlier quoted context omitted.

> And just a reminder, those of you opining based off your experience with GPT3.5... GPT4 is a huge, huge improvement. God, yes. The number of people of HN pushing up their glasses and saying "well, actshually..." when they're basing their opinions off the 3 questions they asked 3.5 is starting to become pretty grating.

The number of people parroting this is also absolutely astounding and grating too. Like, anyone who has spent 5 minutes on this forum already knows this. It’s probably not necessary to keep pointing it out. Yes some people don’t know ChatGPT 3.5 is the default for non-paying customers.

What you are saying is humans are just stochastic parrots - predictable little denial engines regurgitating their latest chosen sentence that makes them feel special and safe.

Any thread in the topic, simple statistical modeling will get you close to perfect to the distribution of arguments that will appear.

Re: Rodney Brooks on GPT-4

#138

Post by winddude »

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…

> 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.

This one. This is the correct resolution.

Re: Rodney Brooks on GPT-4

#139

Post by fwlr »

thomastjeffery wrote: post #107
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…

> 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…

The idea that text contains more information than language is fascinating, wow. Thank you!

Re: Rodney Brooks on GPT-4

#140

Post by p-e-w »

xvilka wrote: post #117
p-e-w wrote: post #82

Earlier quoted context omitted.

I'll let you in on a secret: There aren't actually any "AI experts". There are machine learning experts, that is, people whose expertise lies in designing and analyzing systems that perform (semi-)automatic inference on data. But nobody can be an expert on "artificial intelligence", because we don't know what that word really means. We don't even know what intelligence really is. We have no idea how the human mind wo…

> We have no idea how the human mind works. True, but we already know the so called "neural networks" that many computer scientists believe are how brain works aren't even close. They are all based on half-a-century old concept of neuron that was debunked many times over, experimentally by real neuroscientists.

That's correct, but it doesn't mean artificial neural networks cannot achieve intelligence, or even superintelligence. The fact that the human brain doesn't work like that doesn't automatically imply that (structurally) simpler models are fundamentally less capable.
Post reply on HN