Looks so real !
I think you’d have less dumb ass average Joes cumming over AI if they could understand that regardless as to whether or not the AI wave crashes and burns, the CEOs who’ve pushed for it won’t feel the effects of the crash.
Clair Obscur: Expedition 33
Clair Obscur: Expedition to meet the Dessandre Family
We don’t know how consciousness arises, and digital neural networks seem like decent enough approximations of their biological counterparts to warrant caution. There are huge economic and ethical incentives to deny consciousness in non-humans. We do the same with animals to justify murdering them for our personal benefit.
We cannot know who or what possesses consciousness. We struggle to even define it.digital neural networks seem like decent enough approximations of their biological counterparts to warrant caution
No they don’t. Digital networks don’t act in any way like a electro-chemical meat wad programmed by DNA.
Might as well call a helicopter a hummingbird and insist they could both lay eggs.
We cannot know who or what possesses consciousness.
That’s sophism. You’re functionally asserting that we can’t tell the difference between someone who is alive and someone who is dead
It reminds me of the reaction of the public to 1896 documentary The Arrival of a Train at La Ciotat Station. https://en.wikipedia.org/wiki/L'Arrivée_d'un_train_en_gare_de_La_Ciotat
People used to talk about the idea of uploading your consciousness to a computer to achieve immortality. But nowadays I don’t think anyone would trust it. You could tell me my consciousness was uploaded and show me a version of me that was indistinguishable from myself in every way, but I still wouldn’t believe it experiences or feels anything as I do, even though it claims to do so. Especially if it’s based on an LLM, since they are superficial imitations by design.
You could tell me my consciousness was uploaded and show me a version of me that was indistinguishable from myself in every way
I just don’t think this is a problem in the current stage of technological development. Modern AI is a cute little magic act, but humans (collectively) are very good at piercing the veil and then spreading around the discrepancies they’ve discovered.
You might be fooled for a little while, but eventually your curious monkey brain would start poking around the edges and exposing the flaws. At this point, it would not be a question of whether you can continue to be fooled, but whether you strategically ignore the flaws to preserve the illusion or tear the machine apart in disgust.
I still wouldn’t believe it experiences or feels anything as I do, even though it claims to do so
People have submitted to less. They’ve worshipped statues and paintings and trees and even big rocks, attributing consciousness to all of them.
But Animism is a real escoteric faith. You believe it despite the evidence in front of you, not because of it.
I’m putting my money down on a future where large groups of people believe AIs are more than just human, they’re magical angels and demons.
Thank you for calling it an LLM.
Like a GIF?
Painting?
“LLMs are a blurry JPEG of the web” - unknown (I’ve heard it as an unattributed quote)
I think it originated in this piece by Ted Chiang a couple years ago.
No, not at all.
Remember when passing the Turing Test was like a big deal? And then it happened. And now we have things like this:
Stanford researchers reported that ChatGPT passes the test; they found that ChatGPT-4 “passes a rigorous Turing test, diverging from average human behavior chiefly to be more cooperative”
The best way to differentiate computers to people is we haven’t taught AI to be an asshole all the time. Maybe it’s a good thing they aren’t like us.
Alternative way to phrase it, we don’t train humans to be ego-satiating brown nosers, we train them to be (often poor) judges of character. AI would be just as nice to David Duke as it is to you. Also, “they” is anthropomorphizing LLM AI much more than it deserves, it’s not even a single identity, let alone a set of multiple identities. It is a bundle of hallucinations, loosely tied together by suggestions and patterns taken from stolen data
Sometimes. I feel like LLM technology and it’s relationship with humans is a symptom of how poorly we treat each other.
Ah but have you tried burning a few trillion dollars in front of the painting? That might make a difference!
true
also expecting models to have reasoning instead of the nightmare hallucinations is another fantasy
It’s achieveable if enough alcohol is added to the subject looking at the said painting. And with some exotic chemistry they may even start to taste or hear the colors.
Or boredom and starvation
I can define “LLM”, “a painting”, and “alive”. Those definitions don’t require assumptions or gut feelings. We could easily come up with a set of questions and an answer key that will tell you if a particular thing is an LLM or a painting and whether or not it’s alive.
I’m not aware of any such definition of conscious, nor am I aware of any universal tests of consciousness. Without that definition, it’s like Ebert claiming that, “Video games can never be art”.
Absolutely everything requires assumptions, even our most objective and “laws of the universe” type observations rely on sets of axioms or first principles that must simply be accepted as true-though-unprovable if we are going to get anyplace at all even in math and the hard sciences let alone philosophy or social sciences.
Defining “consciousness” requires much more handwaving and many more assumptions than any of the other three. It requires so much that I claim it’s essentially an undefined term.
With such a vague definition of what “consciousness” is, there’s no logical way to argue that an AI does or does not have it.
Your logic is critically flawed. By your logic you could argue that there is no “logical way to argue a human has consciousness”, because we don’t have a precise enough definition of consciousness. What you wrote is just “I’m 14 and this is deep” territory, not real logic.
In reality, you CAN very easily decide whether AI is conscious or not, even if the exact limit of what you would call “consciousness” can be debated. You wanna know why? Because if you have a basic undersanding of how AI/LLM works, than you know, that in every possible, concievable aspect in regards with consciusness it is basically between your home PC and a plankton. None of which would anybody call conscious, by any definition. Therefore, no matter what vague definition you’d use, current AI/LLM defintiely does NOT have it. Not by a longshot. Maybe in a few decades it could get there. But current models are basically over-hyped thermostat control electronics.
I’m not talking about a precise definition of consciousness, I’m talking about a consistent one. Without a definition, you can’t argue that an AI, a human, a dog, or a squid has consciousness. You can proclaim it, but you can’t back it up.
The problem is that I have more than a basic understanding of how an LLM works. I’ve written NNs from scratch and I know that we model perceptrons after neurons.
Researchers know that there are differences between the two. We can generally eliminate any of those differences (and many research do exactly that). No researcher, scientist, or philosopher can tell you what critical property neurons may have that enable consciousness. Nobody actually knows and people who claim to know are just making stuff up.
I’m not talking about a precise definition of consciousness, I’m talking about a consistent one.
Does not matter, any which way you try to spin it, any imprecise or “inconsistent” definition anybody would want to use, literally EVERYBODY with half a brain will agree that humans DO have consciousness and a rock does not. A squid could be arguable. But LLMs are just a mm above rocks, and lightyears below squids on the ladder towards consciousness.
The problem is that I have more than a basic understanding of how an LLM works. I’ve written NNs from scratch and I know that we model perceptrons after neurons.
Yea. The same way Bburago models real cars. They look somewhat similar, if you close one eye and squint the other and don’t know how far each of them are. But apart from looks, they have NOTHING in common and in NO way offer the same functionality. We don’t even know how many different types of neurons are, let alone be close to replicating each of their functions and operations:
So no, AI/LLMs are absolutely and categorically nowhere near where we could be lamenting about whether they would be conscious or not. Anyone questioning this is a victim of the Dunning-Kruger effect, by having zero clue about how complex brains and neurons are, and how basic, simple and function-lacking current NN technology is in comparison.
I think the reason we can’t define consciousness beyond intuitive or vague descriptions is because it exists outside the realm of physics and science altogether. This in itself makes some people very uncomfortable, because they don’t like thinking about or believing in things they cannot measure or control, but that doesn’t make it any less real.
But yeah, given that an LLM is very much measurable and exists within the physical realm, it’s relatively easy to argue that such technology cannot achieve conscious capability.
This definition of consciousness essentially says that humans have souls and machines don’t. It’s unsatisfying because it just kicks the definition question down the road.
Saying that consciousness exists outside the realm of physics and science is a very strong statement. It claims that none of our normal analysis and measurement tools apply to it. That may be true, but if it is, how can anyone defend the claim that an AI does or does not have it?
This definition of consciousness essentially says that humans have souls and machines don’t.
It does, yes. Fwiw, I don’t think it’s necessarily exclusive to humans though, animals and nature may play a role too.
It’s unsatisfying because it just kicks the definition question down the road.
Sure, but I have an entire philosophy set up to answer the other questions further down the road too 😂 That may still sound unsatisfying, but feel free to follow along: https://philosophyofbalance.com/
It claims that none of our normal analysis and measurement tools apply to it.
I believe that to be true, yes.
That may be true, but if it is, how can anyone defend the claim that an AI does or does not have it?
In my view, machines and AI can never create consciousness, although it’s not ruled out they can become vessels for it. But the consciousness comes from outside the perspective of the machines.
I think this is likely an unsurmountable point of difference.
The problem is that once we eliminate measurability we can’t differentiate between reality and fantasy. We can imagine anything we want and believe in it.
The Philosophy of Balance has “believe in the universal God” as its first core tenant. That makes it more like a religion than a philosophy.
Yeah, I think I see where you’re coming from. It’s a fair point, and we need to be very careful not to loose sight of reality indeed.
The idea of the Universal God is very tolerant towards “fantasy” so far as it exists in the minds of people, yet it also prescribes to align such belief with a scientific understanding. So the thing I’m trying to say is: believe what you want to believe, and so long as it’s a rational and tolerant belief, it’s fine. But it does explicitly recognise there are limits to what science can do for us, so it provides the idea of Universal God as kind of a North Star for those in search, but then it doesn’t really prescribe what this Universal God must look like. I don’t see it as a religious god, but more a path towards a belief in something beyond ourselves.
In the book I also take effort to describe how this relates to Buddhism, Taoism, and Abrahamic religions, and attempt to show how they are all efforts to describe similar concepts, and whether we call this Nature, Tao, or God, doesn’t really matter in the end. So long as we don’t fall into nihilism and believe in something, I believe we can find common ground as a people.
I can understand a desire to find something beyond ourselves but I’m not driven by it.
That’s exactly where Descartes lost me. I was with him on the whole “cogito ergo sum” thing but his insistence that his feelings of a higher being meant that it must exist in real form somewhere made no sense to me.
That’s fair too. I mean, feelings are real, but they are part of a subjective reality that’s not measurable from an objective perspective. But that alone is sufficient to say that science cannot answer all questions, because scientific measurements are inherently limited to objective reality.
Of course there are those that say there must be a single objective reality from which all subjective experiences can be explained, but that’s a huge assumption.
Personally, I think it’s also a dimensional thing. Reality extends beyond the dimensions of time and space, this much has already been scientifically proven. Unless you somehow believe there is a finite limit on the number of dimensions, there will always be dimensions beyond our grasp that we cannot measure or understand (yet).
And bringing it back to the discussion of LLMs, they are inherently limited to a 4-dimensional reality. If those dimensions are sufficient to create consciousness, my position would be that it’s a very limited form of consciousness.
I think the reason we can’t define consciousness beyond intuitive or vague descriptions is because it exists outside the realm of physics and science altogether. This in itself makes some people very uncomfortable, because they don’t like thinking about or believing in things they cannot measure or control, but that doesn’t make it any less real.
I’ve always had the opposite take. I think that we’ll eventually discover that consciousness is so explainable within the realm of physics that our understanding of how it works will make people very uncomfortable… because it will completely invalidate all of the things we’ve always thought made us “special”, like a notion of free will.
If you haven’t watched it yet you’d probably enjoy Westworld - it plays a lot with that space and approaches some very interesting philosophy when it comes to human consciousness and what it means to even be a person.
I don’t know if we’ll ever define consciousness or if we’ll ever discover what it is.
My central claim is that if we don’t do that we can’t convincingly claim that an AI is or is not conscious. We can conjecture about it either way and either guess may be right, but we won’t be able to move past guesses.
I’m sorry, but that article just isn’t very compelling. They seem to be framing the question of “is there free will” as a sort of Pascal’s Wager, which is, umm… certainly a strange choice, and one that doesn’t really justify itself in the end.
The author also makes a few false assertions and just generally seems to misunderstand what the debate over free will is even about.
The example I gave my wife was “expecting General AI from the current LLM models, is like teaching a dog to roll over and expecting that, with a year of intense training, the dog will graduate from law school”
As long as we can’t even define sapience in biological life, where it resides and how it works, it’s pointless to try and apply those terms to AI. We don’t know how natural intelligence works, so using what little we know about it to define something completely different is counterintuitive.
We don’t know what causes gravity, or how it works, either. But you can measure it, define it, and even create a law with a very precise approximation of what would happen when gravity is involved.
I don’t think LLMs will create intelligence, but I don’t think we need to solve everything about human intelligence before having machine intelligence.
Though in the case of consciousness - the fact of there being something it’s like to be - not only don’t we know what causes it or how it works, but we have no way of measuring it either. There’s zero evidence for it in the entire universe outside of our own subjective experience of it.
To be fair there’s zero evidence for anything outside our own subjective experience of it, we’re just kind of running with the assumption that our subjective experience is an accurate representation of reality.
Pointless and maybe a little reckless.
