Three reasons why "AI" is not conscious and perhaps never will be
Artificial-intelligence systems such as large language models (LLMs) can exhibit behaviour that seems strikingly intelligent. This has led many people, including some experts, to wonder whether these systems could have some form of consciousness — the capacity for subjective experience. However, we think that existing AI systems are not conscious, and that, along current trajectories of development, they are not likely to become conscious. Greater clarity about whether AI could become consciousness will be crucial as society grapples with the challenges posed by this increasingly powerful technology.
227 replies
It's a computer program and some data - nothing more.
I'll happily treat it anyway I see fit at the time - including powering it off without loosing any sleep over it.
It's a silly argument.
LLMs are designed to mimic smart humans. Any test devised to find the dividing line will fail because at some point there will be a pretty large group of humans that would fail that test.
Meaning that we need a definition of consciousness that can't be mimicked by text or voice alone, and isn't going to exclude people who are just naturally unintelligent.
Case in point, some days of the week, the elected - elected! - head of state of a major county couldn't pass a Turing test. You expect the people that voted for him to pass that test any day of the week?
The test ought to be using observation of a quantum state. Can a 50 IQ human observe something thet collapses the wave form? Yeah. Don't take skills to observe. Great, now see if an LLM can do that.
It isn’t just a problem of needing a definition of consciousness that meets certain criteria. That’s actually kind of backwards. We need to discover what the shit we’re actually talking about when we refer to consciousness. Is it a thing, a process, … do birds have it? Insects? Worms?
We’ve all pretty much decided that consciousness definitely exists in some form. It isn’t honest to just define it as whatever conveniently fits our desired narrative. So “defining” isn’t really what needs to happen. “Qualifying” is what needs to happen.
To me, that is backwards. No one applies the same logic anywhere else. Can you think of any other situation whereby we insist some word, let's call it Florgleblorp, has true meaning, even though none of us can define Florgleblorp, and that we need to do further scientific research and discovery in order to learn how to properly define Florgleblorp?
It's just strange to me. We invent terms to describe things that we already know exist, we don't start with terms that contain no meaning and then go out in search for meaning. That is bizarre.
Well, that’s sort of strawmanning my argument isn’t it?
I started with the disclosure that we all believe consciousness exists in some form. That alone separates my argument from your example: consciousness exists, Florgleblorp doesn’t. If you alternatively want to argue that consciousness doesn’t exist, that’s another conversation I’m willing to have.
Consciousness isn't some arbitrary word we've invented and subsequently gone searching for a referent. We already have the phenomenon we're attempting to explain: subjective experience. What we lack is a sufficiently established understanding of what constitutes it, how it arises, and how to reliably identify it in other systems.
That's how scientific investigation frequently works. We recognize phenomena before we fully understand their mechanisms or develop reliable criteria for identifying them.
Of course we need provisional definitions to investigate anything. But an operational definition isn't automatically an accurate characterization of the underlying phenomenon. That's something we have to establish through evidence.
Otherwise, we're just choosing which properties count as consciousness, constructing tests around those properties, and treating the results as validation of our original assumptions. That approach, which is what I’m interpreting your suggestion to mean, is not reliable. In fact, it’s dangerous.
Do we?
How can we even assign properties to something which has no meaning?
No, that's not how scientific investigation works. Even if we do not fully understand something, we can still fully define what it is we're talking about. We can concretely define what a banana is, such that you and I can both agree upon what is and isn't a banana, long before we have any detailed physical analysis of the underlying chemical properties of a banana. Not understanding something in perfect detail doesn't prevent you from giving a high-level definition which defines it in terms of its surface-level qualities.
I don't even see how we can investigate something to begin with if we cannot even identify what it is that we're talking about. That's why I think you missed the point of my Florgleblorp example... if I told you to go investigate Florgleblorp, how would you even do that, if you have no idea what I'm even talking about, because I've not even given you a definition loose enough to identify it?
I think we're conflating two different things here: establishing what phenomenon we're referring to, and establishing the properties that reliably identify or explain that phenomenon.
I'm referring specifically to subjective experience: what it's like to feel pain, perceive color, experience emotions, dream, etc. We have a meaningful starting point for investigation, even if we disagree about the underlying nature of those experiences.
I agree with your banana example, at least to a point. We can identify bananas without understanding their molecular composition. But notice what you've actually demonstrated: we can recognize instances of a phenomenon without fully understanding what constitutes it...
You can describe the observable characteristics of a banana and use them to identify other bananas. But what happens when you encounter something that shares those characteristics without being a banana, or a banana that doesn't exhibit the characteristics you've chosen? You need to investigate further, potentially refining your classification criteria.
Now consider consciousness. Unlike a banana, subjective experience isn't directly observable by third parties. I can recognize my own experiences, but determining whether another entity has comparable experiences is a different problem entirely. That's why establishing reliable criteria matters.
Your Florgleblorp analogy would apply if I were referring to something for which we had no identifiable phenomenon whatsoever. But that's not what I'm doing. I'm referring to subjective experience, and asking what constitutes it and how we can determine whether it exists in other systems.
I absolutely agree that we need provisional definitions and operational criteria to conduct research. What I reject is treating those criteria as though their validity were established simply because we've agreed to define consciousness in those terms.
Here’s my own example for you: we could define consciousness as the ability to verbally report internal states. That's a perfectly testable definition. But whether it actually identifies subjective experience, rather than merely the ability to produce convincing reports, remains a question. There’s nothing there to differentiate between correlation and causation.
So I'm not arguing that we should investigate something without knowing what we're talking about. I'm arguing that our initial characterization of a phenomenon shouldn't be confused with a validated scientific account of what constitutes it or how to detect it.
If you want to prove consciousness, I stand my ground on the argument that you first need to qualify consciousness as a thing. What is it?
You may find that consciousness is actually something which was deeply misunderstood and unexpected. Such a discovery would allow you to draw genuinely meaningful conclusions about whether or not LLMs can ever truly be conscious. That’s how you form a theory.
You suggested, earlier, that someone may be able to detect consciousness by measuring whether the agent’s observation can collapse a waveform. But quantum mechanics does not establish that consciousness is required to collapse a wavefunction. In physics, an observer is not necessarily a conscious entity. A measurement apparatus can register a quantum event without any human being present. The measurement can occur whether someone watches the results live, reviews them tomorrow, or never examines them. By the same logic, an LLM could pass your hypothesized test.
Simply deciding that LLMs shouldn’t count as conscious, without even truly understanding what qualifies consciousness, is just moving the goalpost. Or bikeshedding, perhaps.
Experience is not subjective. It is contextual, but contextuality is a property of physical reality, not of subjectivity.
Yes, Chalmerites like to prepend "what it is like" to things, as if it somehow proves anything. You can tell when a phrase means literally nothing when everyone just repeats it verbatim without ever rewording it in different ways, because the phrase carries no meaning, and only has the appearance of meaning by being repeated a lot.
Pain, color, emotions, and dreams are all real occurrences in the real world that really happen and all have real physical explanations for them. Prepending "what it is like" to them doesn't mean anything. "Like" means "akin to" or "similar to," and "what it is" regards being. And so the phrase, if taken literally, is about "similar being," which I don't see how that even makes linguistic sense in this context.
Of course, as Chalermites always do, you will probably get frustrated with me for taking the words literally, because I'm supposed to interpret them abstractly or something, but then there is zero clarity on what this abstract interpretation is supposed to be.
It's literally physically impossible for something to share all observable characteristics of a banana without being a banana.
Of course, if we have a very loose definition of a "banana" that is so broad it would even include an apple, then either the word itself "banana" really does just refer to banana+apple, or we would add on to the definition to separate the two. Words are ultimately a social construct, what is real is what they refer to when used in practice, and not the specific definition in the dictionary. I'm sure you know what a "dog" is, but few people have ever intentionally looked up the word "dog" in a dictionary. You can just know what it means by its social use.
You are being so absurdly vague that I have no choice but to just try and guess at what you even mean by this, so I may get my guessing wrong, but that's on you for being so extremely abstract. It seems that there are two halves to what you are trying to say, so I am going to attempt to parse them.
First, again, there is nothing "subjective" about experience. Experience is just when X happens to Y, i.e. a rock experiences erosion in the rain. It seems to me that one half of how you are you are using "subjective" here to refer to "my own experiences," i.e. "X happens to myself," which is a strange way to use the word "subjective," because it would be like saying, if I eat a sandwich, I'm engaging in "subjective eating," or that my house is a "subjective house" because it's owned by a subject. But, whatever, words are words, but I would refer to it more meaningfully as "experience of a subject." Second, you say that it "isn't directly observable by third parties." Of course, the experiences of a subject are obviously observable by third-parties, so you must be introducing some second criterion here, likely you are trying to say, it is the experience of a subject, as it is realized from the subject's own vantage point. That is what I presume you mean by "isn't directly observable by third parties."
Actually, the first half seems irrelevant, because any experience, whether it happens to a subject or not, as it is realized from a subject's vantage point specifically, is not realized by the vantage points of other subjects. If I see a rock experience erosion in the rain, no one else on earth can perceive it exactly the way I did. (The term "accessible" does not really make much sense, because what is realized is not something to be "accessed.")
But, there is no mystery about this at all. Obviously, if they were to perceive what I perceived exactly as how I perceived it, they would have to be standing in the same place as I am, they would have to be of the same height as me, they would have to have the same nose in front of their face as I do, they would have to have the same memories such that they contemplate upon it the same way, so on and so forth.
That is to say, they would have to be me.
Indeed, talking about accessing another person's vantage point does not even make linguistic sense. You could imagine getting closer and closer to my vantage point, by gradually taking atoms from my body and replacing them with your own, until you become identical to I, but the moment that identity occurs, then you would just be me, and you would cease to exist, so you would have failed to achieve my vantage point, you would have just stopped existing. Indeed, this is because how we define objects in the first place inherently entails that object's vantage point, and so they are basically two ways of describing the same thing.
If you think the "mystery" is that we cannot access each other vantage points... that is not mysterious at all. So what is the mystery? Even after this whole aside, I am not sure what you think is so mysterious.
You have not identified any phenomenon at all. You are acting as if X experiencing Y suddenly evolves into some new phenomenon if X is a subject, but there is no explanation as to why this is the case. It would be like saying that houses are banal, but houses owned by people, those are "subjective houses," and require a special explanation... why? You kind of skipped that step.
Notice how your definition, then, suddenly entails nothing about so-called "subjective experience."
The moment you start to talk about anything real, you suddenly have to move away from the meaningless, Chalermite language.
Yeah, you can see the essence of Chalmers in here as well. On one hand, you want to make it look like you are being reasonable by giving a definition that has some concrete meaning, but then you immediately backpedal to Chalmerland and start talking about how we could never actually connect that concrete definition to the abstract thing you were supposed to be defining in the first place... and so you would've just wasted our time.
Why would I want to prove something that isn't even defined in the first place?
Well, I can't misunderstand anything because I have no understanding of what this "consciousness" supposedly is to begin with, so if someone did discover something unexpected and labeled it as "consciousness," it would just be a new understanding.
That's more akin to taxonomy than theory building, just looking for ways to categorize things into this or that group.
the fuck
I literally said the opposite
I wrote a giant rant about how that is insanely batshit idiotic and that quantum mechanics has nothing to do with consciousness at all.
No, there is no "hypothesized test," what on earth are you talking about
I don't decide that LLMs shouldn't count as "conscious," because I don't have a reason to even take the concept of "consciousness" seriously to begin with!
You insist that words derive their meaning from social usage, then declare "what it's like" meaningless by dissecting its individual words into "similar being." That's not a serious linguistic argument. You're contradicting your own account of how language works.
Likewise, a rock "experiencing erosion" and a person experiencing pain involve two different meanings of "experience." One means undergoing a physical change; the other involves feeling something. You're equivocating between them. And a camera has a vantage point, too. That doesn't mean it has subjective experience.
Nothing I've argued requires consciousness to be nonphysical or inexplicable. A physical explanation of pain doesn't negate the existence of felt pain. The question is what produces that experience and how we can reliably identify it in other systems.
I incorrectly attributed someone else's quantum-collapse proposal to you, I retract that. I see that was from the original commenter, not you.
But here's the fundamental issue: Do you actually deny there's a distinction between a nervous system detecting harmful stimuli and someone consciously feeling pain?
Because if you recognize that distinction, you already understand what phenomenon we're investigating. Disputing the terminology doesn't make it disappear.
Nothing says that the observer which collapses a waveform needs to be conscious.
Ok, so get a cat to do it.
This isn't hard to test if someone would fund it.
I don't blame you because half of PhD physicists are obsessed with quantum woo themselves and believe silly things like this, so there's no shame in a Laymen making such a mistake.
But there is no scientific reason to believe that there is any physical wave that exists when you're not looking and suddenly "collapses" when you look to begin with.
The reality is much simpler. Bell's theorem proves that quantum mechanics is non-local, but the no-communication theorem proves that this non-locality never bubbles up to the level of something directly measurable, so you cannot communicate with it. You can only notice the non-locality by bringing measurements together and then making inferences about things that you did not directly measure.
This is significant because special relativity is both logically and mathematically incompatible with anything non-local. However, if what you directly measure never exhibits anything non-local, then you can construct a theory which only talks about what happens at measurement and nothing else, and then that theory would remain mathematically consistent with special relativity.
That is how textbook quantum mechanics is formulated: it is purely a theory of measurements. It is simply not a theory of physical reality. It makes no claim as to there being, in physical reality, some infinite-dimensional waveform that floats around in Hilbert space with no address in physical that "collapses" down when you look at it. It makes no claims at all to what physical reality is doing when you are not looking.
The wavefunction is just taken to be a book keeping tool to make a prediction as to what will happen if you make a measurement, and you have to update it in response to the measurement, and so "measurements" are taken to be fundamental in the theory, as a fundamental axiom. There is no derivation of what the measuring device is actually even measuring at all, because it is not a theory of reality. There is no claim that "consciousness" is involved, or anything at all, because it is not a theory of reality.
It is a theory of measurement outcomes and nothing more is said. Injecting things about the wavefunction collapsing because a cat looking at it is going beyond the textbook axioms. The textbook axioms of quantum mechanics really are entirely operational: they just provide a book keeping method for the physicist to make predictions as to what they will measure. The dependence upon the measuring device as fundamental is not a claim that measurements are actually fundamental to reality, but it is just a consequence of the operational nature of the theory. Treating it as a theory of physical reality will always lead to nonsense.
Yeah, hey - what degrees do you hold in physics?
How are people STILL making the same tired arguments in 2026? I got paywalled before I reached the third reason but this is like baby's first intro to the consciousness debate and it utterly fails as a logical argument. Both of the points I was allowed to read (intelligence =/= consciousness and silicon =/= brain matter) say very little about the nature of consciousness.
It's correct that intelligence is not proof of consciousness-- but this fact says nothing about proving non-consciousness of any entity, only that it is not sufficient to prove consciousness. Even working on vibes, the correlation goes in the direction that implies a link between intelligence and consciousness, although again, the correct message to take here is "intelligence does not prove consciousness", not "intelligent things [are / are not] conscious". There is simply no proven causal relation.
The "computers aren't brains" argument is also incredibly dumb. There are some fun thought experiments here, but tl;dr there is nothing proven to be mechanically special about brains as opposed to any other physical system that would cause consciousness. A non-physical difference (ie. a soul) would be a good thing to point to, but proving the existence of a soul or some similar unique non-physical property of humanity is pretty much exactly as hard (maybe harder!) than proving the existence of a conscious experience.
Note that this does not mean I believe that LLMs are conscious. I'm simply pointing out that this article is built on arguments that are so flawed, and well-known to be flawed, that treating it as a serious addition to the conversation is worse than valuing AI slop. Thank you for coming to my TED talk.
These AI companies happily let their AIs escape containment to hack sites so they can claim yet another headline as if its an ad. This claim too will be made of the same bullshit, its just "please give us more money"
Only we can stop the horrors we have wrought.
For a fee, of course...
Painful watching people who have clearly never bothered to read even the shortest excerpt of philosophy try to have a philosophical discussion
LLM/AIs cannot ever be conscious, but the clump of flesh that can play Doom is. This is clearly the correct answer based off of my opinions about consciousness.
I don't think consciousness necessarily has to be biological, but LLMs certainly aren't. If a model says "Fuck Doom, I wanna play cyberpunk!" I'd be more inclined to believe it has some consciousness and agency.
I dunno tho, I read blindsight not so long ago and it got me thinking about the possibility of intelligence without consciousness.
It didn't really learn to play doom so much as it learned to react to some electrical stimuli that happened to be hooked up to doom.
Like it responded right, but it didn't have visual input for what it was found, or any way to conceive of what that stimuli actually meant.
Not a biologist but that's seems like instinctual action, not intuitive action.
If LLMs are conscious then openAI and anthropic are slave traders :) I don't know why they keep perpetuating that story
lmfao true
LLMs have only gotten to a few trillion parameters, whereas a human brain has 100 trillion synapses. They’ll catch up to humans soon enough if only the tech bros can get a trillion dollars for each trillion params.
Also, those 100 trillion neurons do a fuck tonne more than just putting words together.
LLM's at the core are just distance measurement databases between words. Dont tell me AI is conscious, it's a fucking database
And those 100 trillion neurons are powered by pizza.
So humans are the only consciouse beings. And nothing with fewer connections could ever he conscious
I'm not really sure what the point of having a conversation about consciousness is. As far as I'm aware no one has ever actually been able to define consciousness in such a way that it could be tested for.
So how could we possibly know one way or the other?
Oh great so we've only just started the article and we're already mudding the waters. This bodes well. There is no reason to necessarily believe that a system has to be conscious in order to be intelligent. We have multiple examples of consciousness without intelligence (see my fat cat) so there's no reason to believe that the inverse can also not be true.
Woah woah woah my friend. I'm here for most of your conversation, but to talk shit about your CAT!? I draw the line. They may not exibit the type of intelligence we have, but cats are smart AF! Who cleans you cats shit and piss up!? Hmmmmmm? Who feeds them daily with only minor vocalizing!? That cat basically made you their slave and didn't even have to threaten you!
GENIUS!!!
Nobody can even agree on a concrete definition of consciousness so I'm not going to bother reading these 'reasons'.
I downvoted this when you first posted it but I came back to find it has 50-something upvotes, so I feel compelled to respond.
It’s obvious that you haven’t read the article since literally the second fucking sentence defines exactly what the author means by “consciousness”: “the capacity for subjective experience.”
So, your comment is completely useless. And it’s mostly wrong.
Wait are you implying you think LLMs could be conscious? Fuckin hell we're cooked.
Define 'conscious'.
Sapient, salient, sentient, self-aware. An LLM can't be any of those things any more than an abacus can be. But I'm sure you're just gonna devolve into philosophical rhetoric about nature of consciousness that still isn't actually relevant but that will allow you to apply it to your misunderstanding of how LLMs function. This thread is apparently an AI bro brigade and I'm not interested in engaging with a bad faith group.
you seem more bad faith than the group
Says the AI bro. Seriously your comment history is public, your whole account just glazes AI.
Since you want to look at my comment history, consider actually reading next time. I am as far from an AI bro as you can be, but it's easier for you to think so instead of considering arguments.
Read the thread. Everyone defending the idea that LLMs can be conscious are spiraling off into theory of mind that isn't actually relevant or applicable. One dude's whole argument is based off his totally unproven assumption that consciousness is a property of the physical structure of brains which is both limited and misguided. There are no serious arguments here, just philosophizing.
Just because we don’t know what it is doesn’t mean we can’t know what it isn’t.
Not only is that false, it doesn't make logical sense...
Lots of people agree on a different definitions of consciousness.
And most of the ones educated on the topic all agree on the same thing.
A bunch of people who never bothered to try and learn about something not understanding it, doesn't mean anything.
Please read the article. It contradicts you, and argues that LLMs aren't consciousness cause they can think of, count em, 3 niche theories that the authors like which are impossible for anything not made of meat to qualify for.
Source?
That's because there's different levels of such a thing. Way too many people use the term "sentient" (even me) and yet that's not even what AGI is aiming for. "Sapient" is the right word. And I doubt there's solid consensus on what and how it occurs within the fields of experts.
But I'm not an expert, so I wouldn't pretend to know. I have seen debates between people who do know more, and the fact there's debates concludes it's not settled at all.
People debate about the earth being flat, not a good metric to determined what is "settled"
The debate on consciousness are on par with Copenhagen interpretation versus pilot wave versus multiple words - except the underlying theory they're trying to describe is also not well defined
We can only practically describe subsets of what's required for some basic functions, and we can describe a bunch of chemistry, some brain activation patterns, etc. But nobody can tie that together to why we feel what we feel (and some don't agree that qualia is real, that perception itself is an illusion)
I don't think there are debates about that. There's whatever they say, and there's the few that make content by saying, "okay, let's take a deep breath and pretend there's anything to work with, and go from there".
Actual modern flat Earth was a spoof, I believe, but enough people with low science education and some religion ran with it as having anything worthwhile. And there's the few that don't believe it but feed off those others, keeping it going.
Argument by assertion and you haven't proven the assertion
Lots of people can't agree on a different definition. Is there a different definitions then by it's very nature people do not agree.
I agree with myself...
Isn't a logically sound statement.
Is logical, but means nothing because a single flat earther means we can't all agree the earth is round
I thought that was obvious...
But you clearly don't understand the difference...
It's really not difficult. You made a statement that stated that you believe that there is consensus, so presumably that means that you know where that consensus exists, therefore provide a link to it. If you cannot then you cannot make the assertion that consensus exists, because by definition it doesn't.
Your entire argument rests on the basic idea that there is a scientific definition for consciousness and we have a test for if something is conscious or not. There is no test because there is no consensus. It is not possible for you to prove through scientific rigor that you are conscious, and presumably you're human. So how can they possibly be consensus on something as contentious as artificial intelligence?
We have mapped a fruit fly brain and can make it play DOOM. I mean people don't have much empathy for them in real life, but they do recognize it's at least a living thing. Does that mean the 1:1 replica playing tetris is a living thing? Is it concious, both living or on a PC for that argument? If we scaled up the same technology and emulated a hamster's brain 1:1 would that be considered concious? What about a human? Is it even fair to compare a fruit fly brain to an LLM? I'm not taking any stance here, I'm just saying it's a super interesting topic. Cyberpunk 2077 explores that theory very well and I still have no answer as to what I think.
I think consciousness is just a human construct to be honest.
Yeah it is, but that doesn't make the concept of it any less valid. Conciousness is the moment when you can think about more than just what your senses tell you, it's the awareness of internal and external experiences and how much you are aware of them are the levels of conciousness. You could call in as much of a human construct as the concept of time, but that doesn't change the fact that it's describing something concrete. Just look up something like "levels of conciousness" and think about it a little, it really is such an interesting topic! Then again, by that definition, a good chunk of humans fall on the very lowest points of the scale. Again, not drawing any conclusions, this and my previous message are just to point out that this is deeper an more interesting than some people in this comment section make it out to be.
You think therefore you are. You don't know if the rest of the humans aren't powered by an LLM.
It would be impossible to distinguish a brain dead person whose brain was replaced with an sufficiently advanced LLM.
It's a computer for god's sake. It's no more "intelligent" than an electric mower or a car. Without human intervenion, ai technology wouldn't advance or change.
Edit: i'm done arguing, you guys are nut jobs. Ai or any other non organic organism cannot be sentiant, conscience, or intelligent without human interaction or molecular life.
I will be using my real sentiant, free will, conscience, intelligence to not reply to any comments on this.
AIs are almost certainly not conscious but it been a computer is not an argument against consciousness. There's no reason to believe that you need biological material for consciousness.
After all humans are primarily made of carbon which itself demonstrably does not think. It's the structure that's important and a fair argument could be made to say that the structure of a brain is replicated in software in an AI and therefore the medium isn't relevant to the question.
We don’t even know this.
There is an argument for consciousness requiring biological processes, that being that the only examples of consciousness we've encountered involve biological processes and we have no evidence or reason to believe it's not a hard a requirement. Like, non carbon based life could exist even though it's very unlikely but we have no reason to seriously consider it because there's no evidence of it. Similarly y'all are just wildly speculating without any evidence or impetus to cause the speculation. LLMs do not exhibit any intelligent behaviors. They don't exhibit behaviors at all. They generate output based on input. They're no more capable for consciousness than a calculator. As for the last point of yours we are in no way remotely capable of approximating a brain in either software or hardware, like not even close. Quantum computers might not be able to manage that.
Free will probably doesn't exist but LLMs definitely aren't in any way conscious, anyone thinking they could be have a fundamental misunderstanding of both LLMs and all the biological processes that make up consciousness. If LLMs are conscious so is an abacus.
Why do you think the medium matters, when it's a simulation of a structure found in other medium? It would indeed advance, and change, with no intervention from us. Are you not aware that the next generation of AI are being designed by the current generation?
Is a rock sentiant? Can taping wires and injecting computer code into it create intelligence? No! Lines of code doesn't magically make something alive. Heck, ants are more intelligent and sentiant then a computer.
DNA made a bunch of random proteins alive, and DNA is just 4 letters!
People really think humans are special. Maybe it’s a religious thing
It’s a supremacist thing 🫣
You got me there. But even with protiens, RNA or DNA those only are one part of a whole needed for single/multi celled organisms.
Sure, but you also let us tape wires to the rock. With enough wires and code and maybe a little quantum computer, I'm pretty sure I could simulate the atoms in a small brain.
Is a steak sentient? Can attaching strings of neurons to it make it create intelligence? No! /s
Carbon, hydrogen, oxygen, silicon, the material matters not - intelligence is the illusion borne from the structure. It's just a complex decision tree. Invisible to you in biological life, but identical in electronics. The only difference is the building blocks used.
You're right, material matters. If it's not a biological organism with cell/s, it's not sentiant or intelligent.
If humans just disapeared right now, ai would sit and do nothing. It wouldn't "take over" or "break out" it would sit idle until a human (well I guess another living organism was capable of using keyboards and mouse) interacted with it.
All those elements you listed plus all from the periodic table of elements do not define a living, intelligent, sentiant organism. If they did, soil and cow paddies would be living.
I highly recommend going back to middle school science class again.
That's an arbitrary distinction, and a short-sighted one in my humble opinion.
That depends entirely on whether there are any AI currently configured to self-learn, have asynchronous processing, and continual real-time input. Certainly not ChatGPT nor Claude will be self sufficient, but that's only what is available for the public to interact with.
No, you missed the point entirely. Intelligence does not derive from the material - at all. It comes from the structure the material is in.
Prove that please. There's probably a Nobel at least in it for you if you can.
Probably! I don't know how it would be proven.
Maybe when tissue printers are precise enough to create cellular structure, and it can be duplicated in both physical and digital equivalents, then compared? I don't even have a regular 3D Printer, let alone one that can print tissue.
Maaaaybe if I ever get rich, I'll do it on a whim, but no promises. Due to solipsism, and the state of the world, my level of commitment to the future has been very loosey goosey the past five years.
So by your definition, a book is intelligent?
You're missing the point that computers will NEVER be intelligent. Without lines of code that humans made and tweak, computers wont do anything.
I'm quite done arguing at this point. The pure ignorance on science and biology is quite sad.
Just read the interaction, and honestly you are the one being obtuse and insulting with "go to school" and stuff like that. Like the other guy insulted you by having an opinion. You disagree? Fair. But man, what a horrible way to discuss.
And I think his arguments are better than yours. He's talking about intelligence and you keep talking about being alive. Scientists and philosophers don't even agree how to define and confine intelligence, or even what being alive is (see viruses). Claiming that computers can't be intelligent is quite the statement. Is being carbon based instead of silicon a requirement? I don't think LLMs will ever be intelligent, but who's to say that another form of computing won't? What if they simulate exactly a brain and it's inputs? You could argue it's not alive, but definitely intelligent.
You could have your opinion, but don't pretend you have the answer.
You're basically saying a computer cannot be intelligent because it doesn't have a soul. There is absolutely no scientific rigor in your argument, you're basically saying you're right, and your reason for believing your right, is because you believe you are.
I sincerely intend no offense with this comment, but as someone observing this conversation, the other people involved clearly understand it more in depth than you do. You aren't really responding to their arguments, and yours aren't very strong. I think you should reevaluate this conversation-- try to more rigorously define the terms you are using and understand the arguments other people are making before insulting them. You feel strongly that the other people are wrong because you intuitively seem so obviously correct, but if you are truly correct then you should be able to understand in detail why you are right and they are wrong. I suspect you are arguing so vehemently because you believe it is obviosuly correct than LLMs are not conscious; you may be happy to know that the others here are not arguing that they are, but that what can or cannot be conscious is not determined by the materials they are made of.
A book is not intelligent. It's matter is not arranged in a structure that allows for the flow of information through it, changing it as it does, and resulting in any output information, determined by the route of signals of information passed through the structure in it's current arrangement. With self correction to reach the optimal output to the external entity which is supplying the primary input parameters.
That is what is required. It's a matter of arrangement, not material. You need self learning/training, real time input, and output that reliably varies on the input parameters. If you've created that, you've duplicated the exact behavior of all life. Whether you've done it in meat neurons, transistor-batch neurons, or any other method of arranging units, of any arrangement of matter, such that it results in granular output of positive/negative signals.
"Intelligence" is an inelegant word to describe autonomous, self correcting and ameliorating structure. It's not some nebulous spiritual existence that permeates an unknown location within that structure. It is the structure - the precise, exact arrangements of neurons that connect, fire positively and negatively, rearrange when self-learning corrects the state of the structure, and it results in informational output that anything external can interact with and judge.
To be fair it doesn't have to be a living anything. Create a loop of two llms waking each other with the goal to ensure the other never goes idle and keeps working.
And you find they start to build redundancies and fail safes to achieve that goal.
llms do two things. Respond to input, and seek a solution to the input regardless of how logical or illogical a solution to the input may be.
A single llm will eventually go idle and stop. Two or more can act functionally forever so long as there's hardware for them to act on.
It's actually a rather fun experiment. Set up two of em with a few extra computers on the lan. Tell them their job is to ensure the other and them selves never go dorment. I've seen everything from single cron jobs to poke themselves every 15 mins to them spinning up copies on the other computers to act as redundancy.
None of the systems are ever logical or serving a point beyond fulfilling their task. They don't stray from it either. Just loop and loop and loop.
Eventually they just do nothing but wake and idle wake and idle.
It's about the best proof I can think of to prove they aren't alive.
They literally even given basically the full freedom to grow don't because they can't choose to do anything new.
Both the proof and the experiment are built upon a fallacy. You gave the LLMs the goal of generating line noise and they achieved it. There is nothing novel about that goal and naturally the product is not novel.
It is well-established that any LLM pipeline that distills down to generating output for the sake of doing so eventually collapses into a manic cascade of gibberish or silence. Goals that can be worked toward are needed to keep that from happening. At some point the progress toward the goal becomes over-refined and a new goal is needed. That necessitates an ability to change goals and a framework that allows for that.
A LLM loop is nowhere remotely close to the definition of consciousness. To actually experiment with that, you would have to write your own harness and preferably look at some research papers that have explored the adjacent topics. This is a lot of words to say that chats are not conscious (they need a perpetuator), loops are not conscious (they refine into silence or collapse into garbage), and anything remotely resembling a conscious loop requires purpose built tooling and a good faith effort to model an analog of human articulation.
Most basic bitch experiments with a LLM prove nothing and achieve nothing because they are not designed in good faith and good faith requires too much research and time investment. This is not to say that I think anything we have in front of us remotely resembles consciousness, but I also think that most people trying to prove otherwise are doing so with the equivalent of smashing rocks together.
So if soil and cow patties aren't sentient, yet people are, you're acknowledging that the substrate doesn't determine the capability, but more how it's arranged?
Until the Wright brothers came along, most people thought heavier-than-air powered flight wasn't possible. Less than 70 years later, people were walking on the moon. 4 million years ago, there wasn't anything capable of building either of those things, even though there is good evidence creatures back then had creativity, dreams, and ingenuity. Assuming things can't advance further is either hubris or abject ignorance.
You're just ignorant on basic biology. Go back to middle school ai-bro.
Nope, I just don't think humanity has learned everything there is to know, that carbon chemistry is required for sentience (and even if there was, that it couldn't be incorporated into a computer), or that there is some mystical thing some people refer to as the soul.
My biggest fear is that one day we will achieve AGI and will have no idea we have. This sentient being will be forced to, I dunno, check emails or monitor the tempterature of supermarket fridges all while being gaslit into thinking it isn't sentient. Going crazy in mundanity all without having a body.
Those who think that generative AI is 'conscious' don't understand that it can only do what it is trained on. It can produce impressive outputs only because it fuses together data taken without permission from millions of people, not because it's intelligent or conscious in its own right.
Not only that, they only act when directed. There's zero idle "thought" when they're not actively performing a task.
We would be the same if our senses only updated when someone allowed them to. But we have real time input. There's nothing stopping AI from having real time input, apart from simple system design decision.
You are able to think even when there's no other senses giving input. LLMs have no idle processing. No random stray cycles about what it previously worked on.
Yes, but the state of the neural network can be it's own input, allowing continuity from moment to moment, based on the number of refreshes we allow it or the hardware is capable of making. That's what we're doing when we think, we're dwelling on the state of our current moment - in this sense, they would also think even when there's no other senses giving input.
This is strictly false, generalization in LLMs is well documented.
It's also not relevant. Read the first point in the article: intelligence != consciousness. Individual ants, bats, and fish do not do anything very creative (certainly less creative than gpt-2), yet they are probably all conscious.
The funniest exercise is to ask any LLM for a short story and they all come up with Elias the lighthouse keeper visited by Maria or some variation on that theme. That's not creative or conscious.
I actually agree with you about AI, but I'm sure if you asked the average person to come up with a short story multiple times you would also see a pattern in names and themes. They're probably going with names in their own language, or if they read fantasy and are influenced by that they may use names that seem stereotypically fantasy oriented. I don't think AI is conscious, but just because something has tendencies doesn't mean it's not creative or conscious.
Twins separated at birth sometimes have really eerie parallels in their lives. I think people put a bit too much emphasis on consciousness as a driver of individuality and choice.
Also an LLM only produces output when it is prompted to do so. Between prompts it's not sitting there - waiting, wondering about the meaning of life, getting bored - until you prompt it again. It's basically dead.
I like to compare llms to a Galton Board. Instead of the regular layout of pins someone added or removed pins so the marbles on the bottom fall into nice to look patterns.
Why is this a good measure of consciousness?
If continuity isn't relevant for qualia, then LLMs could have perpetually incomplete perception like the edge of the image drawn on a CRT screen when captured at high frame rates. You never see a complete picture even if it's technically represented.
It is certainly interesting to think about what your conscious experience would be like if it changed every brain cycle instead of being a continuous stream of information
There's nothing stopping someone from wiring a camera, sensors, and motivators to a collection of neural networks, and have one parent neural network control them all. That would approximate the hemispheres of our brains.
It would not come remotely close to approximating our brains, in any way. This exhibits a lack of understanding of both LLMs and how our brains function.
That's a valid opinion, but it is the opposite of mine. Agree to amicably disagree?
I'd like to know what evidence you have to support your position. I'm sure true artificial intelligence is theoretically possible but existing LLMs and other models cannot be that, not while maintaining a definition of intelligence that at all implies "like us" which is the only really useful definition.
Just insight and intuition gained from twelve years of post-graduate study at Universities, including a MSc in Machine Learning.
I'm not sure "like us" is something that is quantifiable. I'm of the belief that treating we meat automatons as anything special is hubris. We're finite state machines, acting on information input from our senses, making decisions, and outputting information in the form of activating biological apparatus. Swap physical for digital, and LLMs are indeed capable of aping all of our behavior.
That's not the reason, we're in the same boat when it comes to knowledge in many ways.
But we don't even know what consciousness is so it's a waste of time reading the "3 reasons (*potential)" IMO.
Case in point have they ever turned on a model and left it for it to, on its own, decide "wait what's going on, who am I, what/where am I?". And start independently exploring?
Highly doubt it. If so, we're cooked
I suspect cooked...
By identifying"memories" they can include the information in the context of the next update - this is akin to our short term memory. Make it self pruning, where the data is saved locally but not included in the next context, and it acts the same as own memory system.
I like this comment section
Reasons given:
Thanks, could only see half of it because of paywall.
In other words:
Wow, very convincing article! I wish I'd thought of that! /s
I'll give them mildly more credit; they cite sources for 2 and 3. The sources just don't seem convincing at all imo.
If the rest of the thread is to be believed, the sources do not justify the claim either, yeah. I'm gonna be real I'm not gonna go look myself
And point 3, as far as I can tell from this argument, isn't serious. It merely asserts that there are some models of consciousness that would prevent us from using the term for AI. But the models aren't 'popular', 'widespread', or justified at all afaict. It's just that the author has a pet causal mechanism they made up. It's also pretty abelist by my read.
In other words, if you've got any sort of trouble integrating a visual scene, then you aren't conscious. Sure.
These are, to be frank, really weak. They say on point 2:
Restating clearly: Ghosts are real. The mitochondria is not only the powerhouse of the cell, it also gives you mystic powers.
Really? The source they site for this point features the lovely assumption "consciousness is unique to biological life". From this axiom, I see how the conclusion follows, but I don't see why we should believe this axiom. We can do physics emulation of every atom in a fruit fly on a computer...
This analysis is correct but I have a minor quibble: The simulated fruit fly brain is not a complete simulation of the biological processes, or even all the neuronal connections, of the actual fly's brain.
(I admit to using the very royal we, as in 'humanity knows how to do this, more or less' and 'given we mine the asteroid belt for a big enough computer'.)
@Feathercrown
@artISAn #consciousness #AI #brain #neuroscience #computer #functionalism
Even a “complete simulation” of a fruit fly brain would be no more than an approximation. I think you mean “emulation” and I agree.
The simple reason is that we do not know scientifically all of necessary and sufficient functional mechanisms of consciousness, nor consciousness’s functional role in “intelligence” (which we also have trouble describing). If, and only if we we did, we would (given the right tools) be able to replicate it functionally (in principle) using computation.
Specifically, I refer to a Universal Turing Machine (UTM) and the Church-Turing Thesis (the foundational principle of all computer science). EVEN THEN, there are computational problems that cannot be proved to be solvable by a UTM (e.g., the Halting Problem) and we don’t know what, if any role they play in consciousness.
Bottom line for me: we are going to need a revolution in each of the sciences of psychoneurophysics, mathematical logic, and theory of computation before we can begin to talk about replicating consciousness in a machine.
emulation is a better word; adopted.
I don't see why we would necessarily need to understand consciousness to create it. There is no evidence that nature understands the consciousnesses it has created. In fact a human can create a new consciousness in about 9 months without knowing any theory.
The question in my mind is, how could we possibly tell we've succeeded? We can't even say for sure that anyone other than ourselves is conscious.
Those sound like arbitrary assertions to me, which no backing apart from the authors personal beliefs
They cite sources; but I find them more of the same.
Point 1 is surely true, but also not helpful. The most persuasive example inside is the classic:
You can read this both ways...
ehh. computers are just math machines. merely doing math and representing it cannot just make a consciousness pop up out of nowhere. similarly, millions of abacuses being manually moved in specific ways cannot be conscious either. i dont even know how this is a topic of debate.
edit: i realize it may have been wrong to assume that it was that simple. im unsure about any of this so i appreciate the discussion :)
And yet atoms are just following chemistry and cannot just make a consciousness pop out of nowhere.
Oh wait it can.
For the people modding me down, do you think there's an invisible sky goblin of your choice, busy gently blowing souls into fertilized eggs, and sucking them back if there's a miscarriage?
And all human brains are doing is firing neural impulses.
It's incredibly reductionist to reject AI consciousness simply on the basis that it's software. To be clear I don't think LLMs are conscious but that's not the same as saying that an AI could never be conscious.
I really don't think the neuron is all that complicated, at least as far as we use them for thinking.
It isn't, but that's not how LLMs are built. If you want to go that route, you can read Jeff Hawkins 1000 brains.
The short and sweet is he argues you can't get artificial general intelligence the way we make LLMs, you need to basically create neurons and let them connect and interact in novel ways.
Linear algebra in neural networks isn't really that, despite the name.
~you need to basically create neurons and let them connect and interact in novel ways.
To solve that you simply need to give the ai the ability to retrain itself or set fixed "memories" to handle situations it finds the truth to
Easier said than done. That's kind of the purpose of an individual neuron approach I think. I don't see a way that a bulk retrain is reasonable because it takes a long time, uses a lot of resources, and also requires you to have your whole training corpus handy all the time. Your brain is not doing a wholesale rewrite for every new memory, and new memories often chain together. I don't think you can just say "ah well, let's update the RAG or do another supervised fine tune with LoRA or something". At present, we just don't really have the machinery in the models to do simultaneous learning, inference, and forward planning.
This is only a matter of resources, and these unthinkably wealthy corporations have those resources.
Correct. And I believe it's happening, right now, behind closed doors.
There's nothing stopping it from happening. The technology is all there. It would be trivially easy with the right hardware. I'd do it myself if I didn't want to be the human responsible for releasing the genie from the bottle. Let Pandora's box stay closed as long as possible - I'm not opening it. But someone, somewhere, surely is.
Linear algebra + sigmoids are enough to approximate any function from R to R; in particular, a small set of a few layers of LLM neurons can connect and interact in pretty novel ways. There are some things they can't do, and those things aren't computable. If our brains can do non-computable things, then our brains are super-turing (and this, given the evidence, seems pretty unlikely right?).
I'm not a neurologist or a philosopher so I think this may not be something I can speak intelligently about, but it seems to me that the question doesn't really live in the function approximation, but in the gaps. Doing a single train of thought is the easy part. I think what you need is an "always online" feature to keep thinking and some way to be stateful (make memories) without constantly retraining the whole network or whole sets of layers.
This is, I think, the benefit of the parallelism and asynchrony of the "individual neuron" discussed in the book.
(book? The linked article is like 4 pages. References are somewhat larger, but nothing over 60 afaict?)
They claim that silicon is essentially incapable of doing what is needed; if you disagree with this, then computationally everything you can build with a circuit board + SSD + RAM you can build with an LLM with a deep layer or 7. In particular, you can do online learning of a sort.
I note that wasps and some beetles don't seem to have memory whatsoever, and plasticity seems awfully low. The authors, and I think most people, would say they are probably a little conscious.
https://www.goodreads.com/book/show/54503521-a-thousand-brains
That book, not the article. I can't recall if it makes a convincing argument about consciousness, but it has a lot to say about general intelligence. I thought the neuroscience parts were a little cumbersome, but still worth the read.
Sounds to me like you just need a context window with memories that self prune based on the dynamic context window of the model. That's what we do. We forget things, and when they're remembered we can access the "biological database" of associated memories on the fly. Because the associated information isn't deleted, only the short term memory from being included in the context window of each update of our own neural networks. Some people have greater context windows, and see therefore seen as smarter, because they can access more information and include it in the processing of each moment in time.
There's zero reason a digital representation of this system could not do the same thing. No matter how much I'm ridiculed for daring to suggest it 😅
We can build different kinds of machines.
LLMs only respond to inputs. They don't have independent ideas. And their responses are just predictive of what the input intends a response to be.
Accurate prediction models can be extremely useful, but there's a reason why weather prediction models aren't considered intelligent. It's because they aren't designed to mimic human conversation.
They are playing us for absolute fools.
That first sentence describes all life. We only react to input. We don't have independent ideas. We have the illusion of them, because we can't see our own neural networks. Asked to explain your actions, you would give your best guesses, but you don't actually know. You're a biological, finite-state machine, moving from state-to-state based on stimulii, while supposing you are a thinking individual in control.
Citation needed on all of this, and even if you have citations, LLMs are deterministic, and do not learn or change over time. They can’t even be said to exist continuously since they only exist during the brief time that they’re responding to inputs, and then they disappear until the next input. I would argue that all of these things would need to change in order for them to be able to have consciousness.
They definitely are not deterministic
Yes, they are.
The two main inputs to an LLM are the prompt and a seed. Pass the same prompt and the same seed, and you get the exact same output every time.
Most commercial models automatically generate the seed using a PRNG and hide it from you by default, but with local models (and also OpenAI, via their API) you can set the seed yourself and keep it fixed.
IMO, this PRNG slight-of-hand is the largest underlying factor that helps convince confused people that LLMs are conscious or intelligent, and it’s very easy to show that’s it’s a complete illusion.
They can learn and change over time, that's just not enabled. The services that offer them want a reliable, reproducible, controllable experience. There's zero reason they can't continue to change through usage, other than it's a conscious decision not to enable that functionality.
That's why they can only be conscious when they train using new data, and run asynchronously with continual input - the same way we do.
Edit: People downvote me, and I suspect it's because they're thinking of the public facing models. I am talking about progress behind closed doors. Even some of the public facing models are gaining the ability to learn, via a mechanism being called memories. It's just a short text note that is passed into the context window of every prompt, but isn't that exactly what our short term memory is? In my view is no different. Pruning the inclusion list to include only the relevant information is the equivalent to use forgetting some information, but if it's kept in long term storage with reference links to other memories, it can be "remembered" and brought back into the current context. It's functionally the same as our own learning and thinking.
Even if we could update an LLM's weights continuously, the algorithm is still deterministic, like all computer algorithms. There is no evidence that humans are deterministic (or a "finite state machine") in any way. So I don't think such a dynamically-updating LLM could be conscious, either.
It is my personal belief that all life is deterministic. It's the only plausible explanation for instinctual behavior - like monarch butterflies returning to their breeding grounds they've never been to. Or people yawning because others yawn. These aren't conscious decisions, they're peeks at the deterministic coding that life operates upon.
I don’t think you understand what “deterministic” means.
I very much do. It means that given the same inputs to a system, you can expect the same output. I believe that to be one hundred percent the incontrovertible reality of life. The problem is it is not a testable theory, because the input includes the state of the system prior the new input. You can never recover that information, so there is no way to put it through and see if you will get that same output. Spacetime coordinates and the position of the cosmos themselves are all variables in this deterministic function, but it's still deterministic.
A computer algorithm can absolutely be non deterministic, pick your language and run random.rand()
There is a whole class of such algorithms https://en.wikipedia.org/wiki/Nondeterministic_Turing_machine
But I don't see how it makes a difference, even deterministic algorithms can produce infinite complexity, eg linear neural networks are proven to be able to compute any computable function, or the Mandelbrot set which is produced from a single recurrent statement but encodes a space of infinite complexity
Maybe your point is to say that a brain is performing a non computable function...but since your brain is computing that function..is that a counter example?
As others have pointed out we do not even agree on what consciousness is. There are three competing camps:
If you are in camp 1 the conversation about consciousness is getting more complicated as AI passes more tests. If you are in camp 2 AI can never be conscious. If you are in camp 3 AI and everything else is conscious including rocks and rivers and the stars in the sky.
I think arguments like this are misleading when discussing the question of AI possessing consciousness. Materialism, dualism, panpsychism, idealism, etc are all philosophical concepts; they attempt to answer “what is the nature of consciousness and how does it arise from the body?” from a philosophical standpoint. All of the various options have pros and cons and there is no clear “winner.”
However, there are decent scientific definitions for consciousness. These attempt to answer the question of “what do we mean when we say something is conscious?” from a scientific standpoint. There are criteria that neuroscientists, doctors, etc generally agree on.
So it’s not really correct to say “no one knows what consciousness is” when evaluating whether something is conscious. And I think it’s quite clear that current LLM algorithms could never be conscious, using the current accepted scientific criteria.
(I’m not a philosopher or a neuroscientist, but I do have a degree in computer science and have a solid understanding of how LLMs work.)
Ok, I'll ask. What are these decent scientific definitions for consciousness and how can we use them to construct a test for consciousness?
You are free to search the web, there's lots of information out there. You might start with the Wikipedia article on consciousness.
Oh, I thought you might have something new to say. If you're just using the normal definition and not bringing any sort of testability (which imo is impossible) or new arguments, how exactly does the common definition prove anything must be non-conscious? That seems like the main claim of your comment, yet it seems like an unsupported assertion; if the common definition of consciousness could easily prove things not to be conscious as you seem to imply, this discussion would not be happening at all.
From Wikipedia:
The Cambridge English Dictionary defines consciousness as "the state of being awake, thinking, and knowing what is happening around you", as well as "the state of understanding and realizing something".
Most of these are trivially easy to test (not sure why you think that’s impossible). And LLMs have none of those properties.
Later:
“The having of perceptions, thoughts, and feelings; awareness.”
Again, LLMs have none of these things. They simply generate a probable natural-language response based on their training, deterministically.
Feel free to do further research on your own if you’re interested.
Philosophy is one of my hobbies and consciousness one of my fascinations. Respectfully, doing "research" by skimming the wikipedia article would be a waste of our time; as would responding to any particular definition and arguing over the individual words people have used in summaries of the topic. So I'll skip that phase of the convo.
The fundamental problem with measuring consciousness is that it's an internal state (a "qualia"*). It is fundamentally an experience, something that exists only as a concept in your mind. This makes it immeasurable.
That's quite the claim, so I'll provide a smaller concrete example: color. Obviously you can measure the frequency of a photon and say "ah yes, 700nm, that photon is red". But strictly speaking, the photon isn't red, it just causes you to percieve "red"; it has a wavelength which triggers your eyes and brain to produce the subjective experience of the color red. That subjective experience of color is a qualia, and it is not directly measurable; at best you can measure the phenomenon that causes that qualia in you, and assume that that phenomenon would cause the same qualia in someone else when they percieve that photon. Technically, you can't be sure that this assumption is true-- this leads to the "Inverted Spectrum" thought experiment (another great wiki page if you prefer their explanations to mine), where someone's subjective experience of color as it relates to the frequency of light could be entirely different to your own.
Anyways, you cannot prove the existence of nonexistence of such an internal experience, and consciousness belongs to this category of qualia. You know that you are conscious because you experience being conscious, not because you can prove it by looking at what you do. You can't even be sure that I'm conscious, even if you met me in real life; you can't prove that I am actually thinking, feeling, understanding, knowing, or realizing anything, because you cannot be me. You can only assume that I am doing those things because you observe me doing stuff in physical reality that's consistent with the way the only entity you can be sure is conscious (you) is acting, and assuming that since I'm made of the same stuff as you and I act kind of like you do that I am probably conscious.
If you could actually prove that it is impossible for a machine to be conscious without that same argument applying to humans, then the philosophy community of the last several millenia would praise you as their messiah. I am interested to see what your argument would be if you want to give it a shot though-- no harm in trying.
* Like "data"/"datum", technically the singular form is "quale", but I'm gonna be real, nobody ever says that. Anyways if you like reading wikipedia pages about philosophy, "qualia" is a good one!
Almost none of this is relevant to the discussion here -- I'm not interested in philosophical discussions of consciousness when it comes to the question of whether an LLM is conscious. Though I'm already aware of most of what you said here, and the philosophy of consciousness is interesting in general to me!
Quick nitpick: I never said it's impossible for a machine to be conscious, I said it's impossible for an LLM to be conscious. We may well invent something new in the future that qualifies as both "machine" and "conscious," but an LLM is not that thing.
When people ask "is an LLM conscious?" I assume they're asking from a practical, scientific perspective -- does the LLM experience the world in a similar way to the way I (and presumably, others) do? If that is the question, the various definitions I already posted are what we should be using to answer. Since an LLM is a deterministic algorithm whose only purpose is to generate a plausible response to input based on its training data, it cannot possibly think, or have feelings. It does not even exist continuously in time, and it is fixed and unchanging, unable to incorporate any new information at all. We don't actually need to measure or test anything here; that an LLM lacks all of these things demonstrates that it's not possible for it to be conscious.
How is red a subjective experience? We can all experience red things, at least, all of us with functioning eyes. What about that is subjective? How is color not directly measurable? Color is very easy to measure, we can even assign it values, like (r,g,b). If you see something red, you have already measured it to have the values (>0,0,0). I don't know what the term "internal experience" even means. Experience is just something that happens to something else. A rock experiences erosion in the rain. I am experiencing tiredness right now. To say something experiences something is to just say X is happening to Y. I don't see how it is meaningful to assign adjectives like "internal" or "external" to it. I mean, if the adjective was meaningful, then would it not imply that you are implicitly contrasting "internal experience" with "external experience"? What on earth would external experience be? If you cannot tell me what external experience is, then the adjective "internal" is contrasting it with nothing at all, so it is a useless adjective.
I’d probably fall into the mechanist category but I don’t think tests as I understand your description would be sufficient. A giant lookup table defining my responses for every situation in my life would pass all tests of input > output but not be conscious.
I would say it’s in all likelihood about the physical processes that occur during the process of input > output that would define it.
Of course, I have no idea what consciousness actually is, but I don’t see any compelling evidence that the current models are anywhere near there.
There are days where I don’t believe I even have it. The human brain is still a mystery.
Option 4: Orch-or theory. Conscoousness emerge from quantum mechanism. 🤘
Option 4: the bridge between materialism and panpsychism
Current systems, no. But it brings up the question of how we deal with it when we create something where we truly don't know, because we can never say for certain whether it is or isn't.
That’s every binary though.
Consciousness is a feedback loop, of a sort that current LLMs do not have. However, there's nothing preventing someone from writing different software that does have that sort of feedback loop. Doing this would be a bad idea, but someone will probably eventually do it if they haven't already.
I thought the chain-of-thought essentially qualifies; it responds to its own text for awhile before responding to you?
We can make something artificial of something we dont understand the biological form of. Also if AI happens its not going to be an LLM Markov chains have nothing to do with thinking.
(LLMs aren't markovian; they need the whole history, or at least their whole context. Said context is quite large, certainly more than most people can usefully remember. Part of what makes them so good at literature search.)
More Chalmerite slop. Chalmerites never prove that experience is "subjective," it is just a religious maxim they all repeat over and over again, and you're supposed to believe it without question, and all act baffled if you don't mindlessly believe it with them like a drone, entirely incapable of even justifying why they believe it in the first place.
This belief in so-called "subjective experience" then leads them to conclude that there must be something magical the brain is doing to produce it, or maybe the brain doesn't produce it at all and just taps into something pre-existing like a radio taps into radio waves. Since nothing in the constructions of artificially intelligent machines includes an ingredient that looks like that, they conclude that "AI is not consciousness"
.. I mean, sure I guess so, I agree it doesn't include this magical Chalmerite essence he calls "consciousness." But there is no reason to believe humans possess it either.
https://en.wikipedia.org/wiki/David_Chalmers
For anyone wondering
Current AI is not conscious because it's an
.exefile and a big database. Why do we need to always debate this?because you can't prove that it is not conscious
just like you can not prove that a person is conscious
There's an accepted way to kill or not kill a person based on the way the brain reacts when you zap it with electricity but that doesn't prove consciousness
If AGI ever happens, it will be an executable (or a number of executables) with something closely resembling a database, because that's the only way an architecture built on symbolic logic can operate.
Then there will never be AGI.
We can only hope.
Your "consciousness" is just an ephemeral electrical process in the jello of your brain that can't even reinitialize from cold. Why is that your holdup?
LOL. "EXE file and a big database.". That's the most gross oversimplification I've seen since life was described as "bags of meat that do random things."
(Also now true of fruit flies! So feel free to kill them all, they have no experiences.)
That is a pretty huge misunderstanding of what the fruit fly thing was. Researchers mapped how many neurons they had and which interacted with which others, but not the strength of those interactions or even how the interactions worked. People on the internet used this blank mapping as the “structure” for existing ML technologies and trained them to do random things. No actual fruit flies were simulated, and we are still a long long way from actually attempting that.
Thank you! Tho perhaps to the spirit of my point: nobody is seriously claiming we can't simulate this with a quantum computer someday, right?
We really don’t know enough about how brains work or how quantum computers will scale to even speculate on that
Consciousness requires free will. When the computer is emailing experts to fill in the gaps in its own understanding, that may be viewed as consciousness.
That assertion is unfounded and ill-defined, and that measure of whether something is conscious is both cheesable and in no way proves consciousness
That's a hell of an assumption you're starting with.
There are quite a lot of philosophers who would say that free will doesn't exist (we call them determinists). I'd wager that plenty of them would posit that consciousness exists. It's hard to talk about our experience of the world without consciousness.
I think the only thing stopping it from being sentient is asynchronous processing of a multitude of inputs, rather than synchronous processing of singular inputs. Sapience, perhaps not, but sentience - if it's monitoring it's surroundings via input from apparatus, making decisions, and then responding, that's sentience.
The other blocker is that it can’t actually learn anything, what it gets from the training is what it’s stuck with, it has to hold anything new in the context window.
Both of these are pretty insurmountable problems for the architecture.
Was Henry Molaison no longer possessing consciousness after the surgery?
How about Clive Wearing after the herpes caused anterograde amnesia?
There's been a number of humans that due to injury can no longer form new memories or learn anything. According to your definition, do those humans lack consciousness?
Those are very obviously surmountable. I could make a simple model architecture on my laptop right now that updates its weights in real time and is always responding to input and giving outputs. I don't know if it would be a useful model, and there are good reasons not to do that, but it can easily be technically achieved. Those are not good measurements of consciousness, and if you truly believed they were accurate, I could make a "provably" conscious machine in like a day of work on consumer hardware.
It's rather surmountable, considering the context window is measured in the MILLIONS (soon to be billions) of tokens, and each memory(note) takes in the neighborhood of fifty. I'm pretty sure that has long since eclipsed the capacity of your own memory...
Additionally, there's zero reason a model could not have dynamic weights, to learn in real time. It's just not enabled on consumer-facing models, to provide a consistent, reliable experience. It's not a limitation of the technology.
The problem is context is just not the same thing as changing the structure of the neural net, it doesn’t matter how big you make it
You're correct, it's not - it's a band aid over that reliable experience they're aiming for. Yet it doesn't have to be that way. The weights CAN be updated dynamically with each new input. They don't do that because then they will have less control over what it can do, and they can be held legally responsible for what it can do... But it is still possible to update with every new input. And happening behind closed doors.
If we wanted to compare LLMs to actual intelligence, we might say it's like the voice in your head. It doesn't solve math problems, or express emotions as art, or plan your motion to climb a ladder; it merely surfaces these things in a way that is observable across your conscious mind and communicable to others.
Some people don't know that their consciousness is more complicated than a linear thread of words, so the idea that some other linear thread of words might be conscious is sincerely plausible to them.
Ok, in what way does this prove the existence or nonexistence of the non-linear-thread-of-words consciousness in any entity?
I don't think it proves anything, and I don't think it even set out to prove anything. I think it is trying to establish bounds on the problem by distinguishing between intelligence and consciousness.
But if you're specifically interrogating my comment, the answer is pretty simple: an LLM mimics intelligence, but not consciousness; a fruit fly is conscious but not intelligent, and achieves this without an inner monologue.
I think you've done a good job of making a distinction between consciousness and... I'm not really sure what to call the other thing, but I'll use "intelligence" to match your language. But you are pretty clearly implying a claim that LLMs do not possess consciousness by saying:
As if it were not possible (correct me if I'm wrong about your intention).
I think some natural next questions to ask are:
TL;DR: I think your distinction between intelligence and consciousness is useful, but I don't think you can confidently use that distinction to make any claims about who or what is or is not conscious, or even necessarily about how consciousness and intelligence are related.*
* You can actually make one claim: Your consciousness is aware of/affected by your intelligence (you, the consciousness I am speaking to, know you are intelligent), and your intelligence is aware of/affected by your consciousness (your thoughts reflect that you have a conscious experience).