Remix.run Logo
Remix clone Hacker News
new | show | ask | jobsGithub
▲ pixl97 2 days ago

>If it can't learn for itself, then it's certainly not AGI.

What does this even mean, exactly? For example if it can filter new information to put into the next version of itself does that qualify? If not, explain exactly why that's the case.

▲HarHarVeryFunny 2 days ago | parent [-]

Ask 100 people what their definition of AGI is and you'll probably get 100 different answers.

My definition is closer to DeepMind's "can do any [computer-based] task that a human can do", and certainly a human intern can learn on the job and do better on day #2 than day #1.

Learning is basically THE hallmark of intelligence. Being able to learn from experience and use that to do better in the future.

If you could somehow take everything the LLM learnt today and "put it into" an update for tomorrow that would be better than nothing, but the intern on the job is probably learning dozens of things per day that stack upon each other, so that's going to be a S-L-O-W learning AI intern !

Leaving aside the technical issue how you could store a days learning, or what that would even mean, you've also got the privacy and merge issues if this were to be done in the cloud without every customer having their own personalized model.

For real continual learning we need to get past gradient descent-based batch training and develop a new learning algorithm so the model itself learns incrementally as it runs (animal-like predict, observe, learn cycle), rather than being reliant on some external alien to come in, equipped with it's entire learning curriculum, and program its weights.

▲pixl97 2 days ago | parent [-]

OK, let's say we get continuous learning today, what does tomorrow look like?

I won't argue that AI can do some learning as context sizes are still terribly small and expensive to iterate.

What does the world like in a week? A month? A year?

How long before these models drift to their own languages? To their own set of morals? To their own alignment?

None of these questions are answered and I'd rather stay at AGI-lite until they are as having a billion agents going off in their own directions seems like a recipe for disaster.

With current models and their general knowledge self learning just seems like you'd have a few ASIs crop up really quick.

▲HarHarVeryFunny 2 days ago | parent | next [-]

Yeah, I'm fascinated by AGI, but not very enthusiastic about it. I've yet to see anyone, notably not even the people trying to build it, articulate a future where it seems to be a net positive for society.

If it's a choice between everyone living on food stamps or reducing the infinite prime gap to 140, then I guess I'll be happy with a prime gap of whatever it was yesterday.

If/when we do eventually build a more animal-like true human level AGI - build a brain, not just a language model, then potentially we'd be able to build in some of the safeguards that millions of years of co-evolution have built into ourselves, but I expect that is still decades away, and then it'd be capable of doing even more jobs!

The best AI outcome I could wish for is where it is extremely heavily regulated, and AI replacing any human job paying under $1M/yr is banned.

▲jaggederest 2 days ago | parent | prev [-]

I also think it's moving the goalpost significantly. If you asked me what AGI looks like in 2024, I would have said "smarter on tasks that can be done via text than the average random human you'd meet in a NY bodega", and by that standard, it's long in the rear view mirror.

Now I think people are asking a different question, which is "better than any human at any task that can be done via text", which imo is superhuman, not artificial general intelligence.

▲HarHarVeryFunny 2 days ago | parent [-]

It's not moving the goalposts - it's just having a different goal in the first place. This has been Shane Legg and Demis Hassabis' definition of AGI since DeepMind was founded, or before (way before OpenAI appeared), and also for me the goal of AI has always been human level (really a human brain, complete with emotions, etc, but we can start with basic learning/prediction).

▲jaggederest 2 days ago | parent [-]

I guess I fundamentally disagree, because on general knowledge they beat any human, and on processing speed they beat any human, and on problem solving likewise. I just don't see a domain besides "retain change over time" that they fail on, and that is something you can do in the system rather than the model - coding agents can and do build their own memory system and use it, like an amnestic human writing notes.

I think what I'm objecting to is the idea that if it can't do every task that every human can do better than any human, it's not AGI - I think we have to allow that this is a fundamentally different form of cognition, so requiring a 100% match feels uncharitable. Every human has cognitive gaps that AI doesn't have.

▲HarHarVeryFunny 2 days ago | parent [-]

I'm not sure there is anything to disagree on here - just different definitions of what we personally associate the term "AGI" with. There is no magic to the name, and however you choose to define it there will be less and more capable levels of AI that precede and follow it.

Separate from what you choose to call any given set of AI capabilities, learning is generally more than retaining change over time. Learning isn't the same as memorization - it's the difference between memorizing a stack of how-to books and going out and practicing the skill for yourself, and all that entails.

A learnt skill needs to be internalized such that in the future it's what is driving behavior, which means it needs to be in the weights. Imagine a robot trying to learn to play tennis where rather than updating it's connections between perception and action it's trying to store everything as a bunch of notes!

It's not just physical tasks where this applies - the same would be true to trying to teach a multimodal model to recognize different types of mushrooms from photos, or even learning a new language. If a system can truly learn for itself then it shouldn't need to be pre-trained with language, but rather should be able to learn by immersion and practice the same way we do, and the same way a dog learns to follow commands. You can't do this by leaving notes to yourself - the learning mechanism needs to be animal-like and update the systems weights in response to prediction failure - a runtime version of the way SGD-based pre-training updates weights in response to prediction errors.

▲jaggederest 2 days ago | parent [-]

> If a system can truly learn for itself then it shouldn't need to be pre-trained with language, but rather should be able to learn by immersion and practice the same way we do, and the same way a dog learns to follow commands. You can't do this by leaving notes to yourself - the learning mechanism needs to be animal-like and update the systems weights in response to prediction failure - a runtime version of the way SGD-based pre-training updates weights in response to prediction errors

I am not sure why any of that is necessarily true. I think you are crediting the human brain with a lot more than it actually does - memory is something in the direction of a fine tune on existing neural hardware, it's really not anything special, and I absolutely think that "just taking notes", when taken to a sufficient fidelity, is enough to pass that bar.

▲HarHarVeryFunny a day ago | parent [-]

You can't take notes before you've learned language, so if that's how you propose to learn then you are doomed to build something with less learning capability than an human brain.

Maybe you don't care - if you just want to push to see how far an LLM can go that's fine. but it's not going to be superhuman-level, or even human-level, if it doesn't have the learning ability of a human. If it relies on language and notes to itself then it is just an LLM.

▲jaggederest a day ago | parent [-]

Okay, nowhere did I say it would be before they learned language. I said we could use models that can trivially build notetaking systems to auto-improve themselves by taking, refining, and updating notes as a memory system to exceed the x token context threshold.

> but it's not going to be superhuman-level, or even human-level, if it doesn't have the learning ability of a human. If it relies on language and notes to itself then it is just an LLM.

Why do you think that? You do a lot of assuming a thing to be true without any clear reasoning or evidence. I'm talking about a system, including a LLM, that acts in a way that is equivalent to or better than an average human at a large range of tasks, such that is generally capable of learning and problem solving similar to the degree of success humans have. All these specific criteria about what is "actually human" is not the point. We know how to make more humans, I'm talking about a form of intelligence that we are creating that is genuinely alien, so holding to "well it doesn't work the same as a human" is an impossible bar. If you reverse the situation, humans are by no means LLM-equivalent intelligences - perfect memory of long form content, rapid calculation, instantaneous tool calling - a hypothetical silicon based lifeform would dismiss us as "really not intelligent at all, they can't even do multiple floating point calculations per second in their head", and I think that would be similarly silly.

▲HarHarVeryFunny a day ago | parent [-]

OK, so we're basically talking about different things. I'm talking about building something that has animal/human intelligence and learning capability, and would be able to first learn language for itself (no pre-trainiing), and then proceed from there.

You're talking about building an "alien" intelligence, apparently something based on an LLM, where language is baked in from the start and can therefore be used as the basis of some type of learning (or at least memorization).

Sure an LLM can be super-human at specifc things like math where the traditional computer strengths of compute, memory, etc apply, but in the context of AGI (generality, not a bag of narrow intelligences) you can't call something super-human if it can't at least also do the majority of things that a human can do, and not having learning ability as powerful as a human seems to me to be a pretty massive omission.

If WE didn't have ability to learn language, then we'd not be here talking about it, and there would be no LLMs. Having a fundamental ability to learn, isn't just useful to learn language from scratch, or for learning all the non-linguistic skills a human is capable of, but would also support an AI that can learn things that we cannot (e.g. could give it additional "senses" like global pressure/temperature inputs so it could "see" the weather and learn global patterns).