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

Many in the AI field think the bigger-is-better approach is running out of road

economist.com

Post reply on HN
Print view View original

281–290 of 354 posts

Re: Many in the AI field think the bigger-is-better approach is running out of road

#281

Post by thfuran »

williamtrask wrote: post #233
roenxi wrote: post #198

Earlier quoted context omitted.

If brains aren't a complex probability machine, how is it possible that people get the same sort of math problems right and wrong in an inconsistent manner? Or mis-speak? It is undeniable that human reasoning is a stochastic process. Otherwise it wouldn't be reasonable for people to make mistakes after learning something. Especially inconsistent mistakes, like when we give people 10,000 addition problems to do in a r…

Food for thought: randomness is in the eye of the beholder. It’s only random to you if you don’t know how to predict it. So can a mind ever truly be stochastic? Perhaps to the minds of others, but never to itself.

No, it's random if it can't be perfectly predicted with perfect knowledge.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#282

Post by nicetryguy »

michaelteter wrote: post #63

Isn't the fundamental problem that LLM's don't actually understand anything (as greater concepts), but rather operate as complex probability machines? My 2 month active experience with ChatGPT-4 gave me the following takeaways: - when it's right, it's amazing; and when you, the operator, can recognize the niche use case where it performs really well, it can be a game-changer (although you could have programmed a tool…

> Isn't the fundamental problem that LLM's don't actually understand anything (as greater concepts), but rather operate as complex probability machines?

I'm genuinely unsure if my own brain is any different.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#283

Post by thfuran »

jiggawatts wrote: post #72
TeMPOraL wrote: post #42

Earlier quoted context omitted.

How much general "thinking"[0] would you want those "tight little specialist models" to retain? I think that cramming "all of the web" is actually crucial for this capability[1], so at least with LLM-style models, you likely can't avoid it. The text in the training data set doesn't encode just the object-level knowledge, but indirectly also higher-level, cross-domain and general concepts; cutting down on the size and…

Even just for spell checking, having a ton of general knowledge helps. Knowing the correct spelling of product names, British vs American English, context-specific alternate spellings, etc… Even GPT 3.5 has trouble following instructions, but I’ve found that GPT 4 is almost flawless. I can tell it the document uses Australian English but to preserve US spelling for product names and it’ll do it! One quirk is that it’…

Can you not just tell it not to muck with proper nouns?

Re: Many in the AI field think the bigger-is-better approach is running out of road

#284

Post by usrbinbash »

p-e-w wrote: post #259
usrbinbash wrote: post #230

Earlier quoted context omitted.

> Do you have any evidence that contradicts that claim? Empirical evidence? Yes I do. The brain commands an entity that has to exist and function in the context of objective reality. Being unable to verify it's internal state against that, would have been negatively selected some time ago, because stating: "I'm sure that rumbling cave bear with those big sharp teeth is a peaceful herbivore" won't change the objective…

That's speculation, not evidence. The traits you describe aren't demonstrably incompatible with the mechanism I proposed.

It's empirical evidence, since we exist and are very much capable of selecting the correct statement from a bunch of stochstically likely, but untruthful statements about objective reality.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#285

Post by usrbinbash »

zztop44 wrote: post #276
usrbinbash wrote: post #230

Earlier quoted context omitted.

> Do you have any evidence that contradicts that claim? Empirical evidence? Yes I do. The brain commands an entity that has to exist and function in the context of objective reality. Being unable to verify it's internal state against that, would have been negatively selected some time ago, because stating: "I'm sure that rumbling cave bear with those big sharp teeth is a peaceful herbivore" won't change the objective…

I think this is incomplete on a number of levels. For a start, to be interesting, “truth” has to be something than just whatever your eyes can see. There have been wars (culture, economic, kinetic, etc) fought to define something as a truth. The concept of truth is notoriously hard for humans to grapple with. How do we know something is true isn’t just a neurobiological question, it’s been grappled with throughout th…

> And for the record, rumbling cave bears are mostly peaceful herbivores.

For the record, all members of the Genus Ursus belong to the Order Carnivora, which literally translates to "Meat Eaters". And that includes Ursus spelaeus, aka. the Cave Bear.

And while it most likely, like many modern bears, was an Omnivore, that "Omni" very much included small, hairless monkey-esque creatures with no natural defenses other than ridiculously small teeth and pathetic excuses for claws, if they happened to stumble into their cave.

> The concept of truth is notoriously hard for humans to grapple with.

I am not talking about the philosophical questions of what truth is as a concept, nor am I talking about the many capabilities of humans to purposefully reshape others perceptions of truth for their own ends.

I am talking about truth as the observable state of the objective reality, aka. the Universe we exist in and interact with. A meter is longer than a centimeter, and boiling water is warmer than frozen water at the same pressure, whether any given philosophy or fabrication agrees with that or not, is irrelevant.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#286

Post by elif »

krisoft wrote: post #268
elif wrote: post #254

After a year of Tesla FSD beta I agree 100%. I used to think that adding more and more and more real world video events to the training pool was contributing at least a diminishing value to the model. But the worst of all behaviors have not diminished and some have actually gotten worse. Most of the improvements now come through what feels like manual heuristics and tuning parameters rather than any actual improvemen…

> Most of the improvements now come through what feels like manual heuristics and tuning parameters rather than any actual improvement in intelligence. So if I understand your are driving a Tesla car with Full Self-Driving (FSD) capability, but you are not an engineer at Tesla who is privy to implementation details. How do you know if a change you perceive is caused by model retraining vs manual heuristics and tuning…

You missed the part of where you quoted me where I said "feels like" and invented a part where I said "have empirical evidence"

Though to answer a softer form of your question which is based in my feelings, I would say through a combination of my experience writing AI's, coursework in my AI specialized computer science degree, and through reading the patch notes which often state explicitly the tunings I mentioned.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#287

Post by brutusborn »

tchaffee wrote: post #176
brutusborn wrote: post #166

Earlier quoted context omitted.

It can be justified in a deterministic universe because it makes criminals less likely to commit crime in future. As a recipient of punitive justice myself, being punished had a tangible effect on how I thought about crime and thus how I behaved post-punishment. Whether you believe that was deterministic or due to my own free will doesn’t change the outcome.

Punishment is not effective for most people in reducing crime. One thing that is effective is increasing the perceived risk of getting caught. People rarely commit crimes when they are sure they will be caught.

I hear this often but don't see much evidence of it. Do you know of any good studies to support the idea that punishment isn't effective?

Re: Many in the AI field think the bigger-is-better approach is running out of road

#288

Post by regularfry »

usrbinbash wrote: post #236
regularfry wrote: post #201

Earlier quoted context omitted.

What's needed, ideally, is a checker. Something that takes the LLM's output, can go back to the training material, and verify the output for consistency with it. I don't think those steps are out of the bounds of possibility, really.

> Something that takes the LLM's output, can go back to the training material, and verify the output for consistency with it. The problem is what you mean when you say "consistency". The LM checks if sequences are stochastically consistent with other sequences in the training data. Within that realm, the sentence: "In the Water Wars of 1999, the Antarctic Coalitions aramada of Hovercraft valiantly faught in the battl…

Yeah, I don't mean stochastically consistent. Semantically consistent. The job of generating content from text and the job of assessing whether two texts represent aligned concepts are two different jobs, and I wouldn't expect a single LLM to do both within itself. That's why you want a second checker.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#289

Post by nathan_compton »

bemmu wrote: post #111
michaelteter wrote: post #63

Isn't the fundamental problem that LLM's don't actually understand anything (as greater concepts), but rather operate as complex probability machines? My 2 month active experience with ChatGPT-4 gave me the following takeaways: - when it's right, it's amazing; and when you, the operator, can recognize the niche use case where it performs really well, it can be a game-changer (although you could have programmed a tool…

I’m not convinced the language part of my brain isn’t just a complex probability machine, just with different trade-offs.

There isn't really, properly, a "language part" of your brain. There are certainly areas of the brain that are more active or participatory in language generation, but the division of the brain into functional areas is much more conventional than, for example, the division of a computer into a mother board, a processor, a GPU, etc. In a much more rigorous sense, these computer parts constitute discrete units. Your brain is a bag of cells for which functional divisions are, in a very real sense, incidental.

To give a very concrete example, the part of your brain called the visual cortex by neuroscientists nevertheless is activated by auditory and other sensory stimuli and assuredly participates in the processing of other sensory input in other ways. I am very suspicious of any attempt to talk about the brain which forgets that it is, in a very real sense, a gestalt.

To your specific point, and taking into account all I've written above, I've got little doubt that some part of your brain really is a probabilistic language generating machine. But the exact point here is that your cognitive abilities constitute much, much more than merely the ability to generate plausible language. Indeed, as I experience complex cognition, conversion into language is often the last and most trivial part of the exercise.

Re: Many in the AI field think the bigger-is-better approach is running out of road

#290

Post by chongli »

MrScruff wrote: post #272
chongli wrote: post #225

Earlier quoted context omitted.

Is my model of the world that different to that of an LLM? Well, one major way you’re different from an LLM is that you’re alive. You’re capable of learning continuously as you go about your day and interact with the world. LLMs are “dead” in the sense that they’re trained once and frozen, to be used from then on in the exact same state of their initial training.

I agree that is a fundamental difference. That’s what I meant about reinforcement learning. Our ‘model weights’ are being updated with new data all the time. I was just referring to what happens at a specific instance in time when someone asks me for example ‘What’s the capital of Norway?’

That one’s not a great example. Either you know the capital or you don’t. There’s no process (other than research) by which you can learn the name while attempting to answer.

A question I get much more often is “how do I solve this math problem?” Many times, the problem is one I’ve never seen before. So in the process of answering the question, I also learn how to solve the problem too.

Post reply on HN