Comment by moezd
1 day ago
Oh come on. Objects are objects because you want to freely observe them without consciousness and privacy and taboo attached to them. Entities are entities because you want to assign credit or blame to them, so you want to believe that they have agency so that you hold them responsible. We did this with sharp volcanic rocks once, calling them evil. Did it help? No, we got idolatry and eons of bigotry until we secularized once again.
Never forget that LLMs are just next token predictors. The harnesses are what's getting uncanny. And credit/blame assignment? Just assign a surrogate for now, let them deal with the consequences ffs.
Here is a sketch of my argument. It would need tidying but that would require a lot of work.
I've been calling LLMs digital intelligence. That's in my opinion what they are and on the digital/conceptual realm they are better than humans since LLMs are better generalists.
But I'd never state their conscious. For me consciousness extends outward from myself, since I can't trust anything else. It starts with the question like: why can't I control the movement of another person? Why can't I transfer my mind to another body and theirs to mine? Why is it that whenever I sleep, and similar things, I end up back to this place called "reality"?
I come to the conclusion that I am conscious and I'm a conscious being in reality. I am human and I've been raised by humans. Now, technically, all humans besides me could be zombies. They could be non-conscious beings that simply can act like humans. I'm sure we'll be able to create them in a few decades. However, I'm incentivized for multiple reasons to believe they're not zombies. Therefore humans are conscious too. All of them. Can I know for sure? No since I can't even trust my own senses fully or my own thoughts (the whole Descartes thing) but I choose to trust those things too because I need some reasonable-ish foundation to work from.
If humans are conscious, why aren't animals like us? Surely great apes must be conscious at least to some extent. Hell, some of them even have better short-term memory than us [1]. Well, turns out, from an anatomy standpoint they look a lot like us. Fine, let's assume they're conscious too.
Now we get into the territory that anything that has a large enough brain must be conscious because that's how we govern our consciousness. So you can extend this to all kinds of living beings.
I did this from my perspective. Of course, you should do it from your perspective. The same argument could be made.
For as long as we can't manipulate conscious experience in some way shape or form, we have no clue whether anything that isn't like us can be conscious. We'd need to be able to merge and split conscious experiences or we'd need to be very strongly able to understand why that isn't possible. And I don't mean just at the biological level but at the experiential level. Currently conjoined twins, and what they tell us gives some insight (being able to sense the other part that's part of the other twin).
Other than that, I claim we have no clue what consciousness is other than that we're experiencing it. So to call LLMs conscious is way too big of a claim. But they are definitely intelligent because they come up with things that I wouldn't have and it's useful to me. Perhaps a practical characterization of intelligence but it works for me.
I think for anyone to say something useful from this you'd need the trifecta of deep neuroscience knowledge, deep philosophy knowledge and (at least) a strong understanding of how LLMs work.
Humans are just baby replicators. The civilizations are what's getting uncanny.
It's good the 'just token predictors' crowd is starting to pivot and hedge though lol. 'It's just the harness!' uh huh. give it a couple more months and you'll be in here regurgitatin' Gordon Wood
Baby replication (creation, actually, I don't think anyone has ever copied a baby) is not what humans do as their only function that enables them to do anything in life.
LLMs are literally token predictors. You can debate over the "just", but we haven't moved onto some new text generation technology, we're just expanding that one. I don't understand why you'd scoff at facts.
I think a better angle is that LLMs are "just" token predictors in the same way that we are "just" made out of atoms. A lot of people do not believe consciousness can be material; they don't believe that you could build conscious beings just out of normal matter, without an immaterial "soul". Whereas if you think that physical matter is a lot, actually, and you can build a nearly infinite variety of things out of these building blocks... well, why not consciousness?
Predicting the next token is an immensely powerful thing. You can encode basically anything into a sequence of tokens, any configuration of matter. If you have a perfect next token predictor, you can predict basically anything. You can simulate an entire universe. It's Turing-complete. It sounds diminutive, but it's a general paradigm. Given everything this paradigm appears to be capable of, well, why not consciousness, really?
1 reply →
Oh so you're saying the biological imperative that underlies all human function, upon which all other emergent functions have sprung from, isn't actually the correct way to characterize human capability and is over-reductive..and strange?
Agreed.
Now just take that one extra logical step please
Baby replication is one thing humans might do.
Word fragment generation is the one-and-only thing llms have ever done.
I think you can similarly reduce humans. For example, all humans do is contract muscles. This doesn't fully cover all of biology, such as replicating cells, etc. so not exactly comprehensive, but hopefully that helps get the idea across.
LLMS: Receive text/binary input, process it, and output text/binary input Humans: Receive various inputs, process them, and output muscle contractions
You're quoting Good Will Hunting... Is there anyone among us willing to get up on a desk and quote "O captain my captain", too? My sweet summer child, just because it maintains your 2000+ dependency filled codebase for you, it doesn't mean it can ever hope to master the mysteries of the universe.