EDIT: I mean neuroscientists know brain frequency is more than an accidental correlation with consciousness.
I mention "unconscious" only to help disambiguate an overloaded word. It's used for both "this model includes something in its representation" as well as "this model is currently executing".
People often just hear the phrase "brain waves" and immediately think of "woo woo spirituality" because all the spiritual crack pots are constantly talking about waves without even having a physics background.
How about picking up a physics textbook? The status quo for Quantum is Griffith's Introduction to Quantum Mechanics. Might I also suggest his E&M book (again, a standard). If you're half as passionate as you portray, I'd also pick up Jackson. That's the standard graduate level E&M book (but a warning on this: to get through Chapter 1 you're going to need to be intimately familiar with PDEs and Green's Function. To get further, I sure hope you like Bessel functions).
Take your passion and use it to drive you to prove. Don't just conjecture and don't try to derive things from first principles. You will get it wrong because you won't actually be working from first principles. Again, don't ignore humanity's greatest invention. It's so much easier to catch up than reinvent the wheel.
Also, you should become familiar with Penrose. I'll leave it to you to figure out why.
Have you done substantial reading?
> because they have no state that represents anything.
The point I'm making is that representation is not enough. Storage (or memory) is not enough even recall is not enough. If it is, then any mechanical process is. Because you can build computation machines out of nearly anything. You can build memory out of almost anything. And the hello world python program I wrote has both!
You can argue scale, but we see animals with much smaller scale than these ML models perform unambiguous reasoning tasks and critical thinking. You can then argue architecture and data, but then that's not just scale.
The problem with all these people making claims is that they don't consider base assumptions and they don't follow the consequences of the model (your mental model of what is consciousness) is and must mean! So many people argue that they are working from "first principles" but I don't see anyone writing down their fucking axioms.
Of course you should try to prove what you want, but not at the cost of not being wrong. Doing so will prevent you from achieving what you want. I'm just so tired of people shooting themselves in the foot and calling it medicine. We need to stop with all the bullshit because I actually want to build AGI
Sure then following OP's model outlined here https://news.ycombinator.com/item?id=41427796, rocks are simply conscious of the passage of time. Perhaps it can be said that is the one thing every matter in the universe is conscious of.
>The point I'm making is that representation is not enough. Storage (or memory) is not enough even recall is not enough.
You don't know what is or isn't enough.
And by the way, what 99% of people think of a concept they can scarcely define, test for or prove doesn't mean shit. We are regularly wrong about concepts with much stronger standings.
Humans don't have any special insight to consciousness. We can't even make decisions without unknowingly altering our preferences and fabricating reasons in the process. What makes the conscious mind tick was clearly lost along the way.
>We need to stop with all the bullshit because I actually want to build AGI
No, you want to build the idea of AGI that exists in your mind. Clearly not Artificial General Intelligence seeing as what exists today checks all those boxes.
You want something more, and that's fine because I do too. Our state-of-the-art models are baby AGI, because they can accomplish a wide variety of tasks without being specifically trained for it. They can do limited reasoning on new tasks. That's intelligence that's both artificial and general. But it's just the start.
There's a few axes that need to be captured. Breadth of knowledge, ability to reason, and ability to update the model based on new information. LLMs are great at 1, possible superhuman. For 2, they're 1/10 on a scale of 0 to human. For 3, fine tuning and prompt stuffing get the job done for now, but they're a far cry from human ability.
And yes, you can build computation machines out of nearly anything. If you simulate a human brain with sticks and stones, what does it matter the substrate?
> AGI exists now
What I hate about these conversations is that people learn the basics but there's been a lot of advancements since the 50's. We've learned a lot since the 90's too! Yes, the metrics keep changing, but they are not in response to the current state of the art in a specific effort to make the systems not AGI[1]. But rather the criteria keeps changing because we keep learning. It would be ludicrous if it didn't! Because WE STILL HAVE NOT CONCRETELY DEFINED CONSCIOUSNESS, REASONING, THINKING, and many other similar terms. The criteria are generally guides created by best guesses. But until we have formal definitions, expect it to change (and when we do have formal definitions, be sure as hell to understand their assumptions!).
But I can confidently tell you that LLMs do not exhibit generalization in the way we've been using it for decades. This is easy to demonstrate but if you're a true believer then there's no convincing. And in that case, just don't bother arguing either.
I really do encourage people to take their passion and use it to drive gaining a deep expertise on these subjects. You're passionate right? I hope that passion is not about "being right" or "having answers" and is actually about "understanding." And if you can't differentiate that, I suggest starting at the beginning, since these topics are covered there.
[0] Sure, you can find examples of people stating criteria, but I can find you samples of idiots, even among scientists.
[1] There are some bad faith people who do this but they're terrible people too. You may freely disregard their opinions.
It's not about animate vs inanimate, it's about models that represent something.
There's also some confusion because we have limited language here. A person can be "unconscious", but that doesn't mean they've lost their world model. It's just currently not executing.