What position are you taking? (Genuine question.) Do you believe humans are conscious because of a special property we have? If you are a dualist, how do you explain the interaction in physical space between the non-physical special property and the physical brain?
The universe is material of course, the soul is made of matter, thoughts are made of matter and so on. The question is what kind of matter? Looking at the brain, we won't find any intelligence in the motion of its cells. They mostly stay in place. If we go one level deeper, we'll see molecules, and their motion reveals something more interesting: various neurochemical reactions. Yet, there is no intelligence at that level either. Going one more level deeper we'll find electrons and that's where things become somewhat intelligent. At the current level of science, we can't see what makes electrons move, but if we could, we would discover finer and finer forms of matter until we would reach an end. That finest level of matter for that particular person we would call his soul. I think it's the presence of those higher imperceptible levels of matter that creates consciousness. If it all stopped at the cellular or molecular level, we would be like complex mechanical watches, not very different from LLMs.
I do. God. I am partial to and bullish on the brain being a quantum-classical computer. There has been interesting research exploring that recently [0, 1, 2, 3]. That (quantum mechanics) seems to be escape-hatch, so to speak, from the otherwise deterministic, mechanical procession of the universe.
For it to be an "escape hatch", there'd need to be a measurably different distribution of outputs than what you'd otherwise expect, that translates to functions that are not Turing computable. There is zero evidence for that.
I agree. I haven't read any more than the first paragraph yet (but I will, after work). The only way that 'AIs are not conscious' can be true is if we decide, with high confidence, that they are lacking some essential property that is not lacking in ourselves. There is no convincing philosophical position that supports this (convincing to me, anyway).
Hmm let me try then. AI agents do not experience real time. This is understandable given their design, but it’s also something we have good empirical evidence for. They cannot, especially over the long horizon, track how much real time has passed as they complete their tasks. And they are not off by a few minutes but often bizarrely off, even mixing across past present and future.
Biology, on the other hand, is nothing but timed processes in a loop, the most obvious to us being the circadian cycle. As estimators of wall clock time, biology isn’t great, but when it comes to internal processes, and most certainly learning, memory, sensing, locomotion… biology is rhythmic in behavior, and the rhythms go all the way down to gene expression. More, these rhythms are, except during sleep, constantly entraining to signals from the environment that indicate time, most importantly light.
I think it’s a fairly unremarkable claim that agency and consciousness are temporal processes that depend on systems having an internal sense of time. How else can you anticipate? How can a system that can be literally turned off ever succeed in an environment where time never stops?
I agree, there is certainly a clear operational difference in how human brains and these models produce language. I'm not convinced, though, that experiencing real time is a requirement of conscious thought. A cognitive system (assuming these models and the brain are both examples of this) needs to proceed from state A to state B as it processes information. The time interval between these states can be arbitrary, it seems to me (setting aside problems with disconnecting the mind from its substrate, which does of course need rhythms at various frequencies; oxygen at a higher frequency than sugar, and so on). I'm not sure that going from A to B quickly, or slowly, or with intervals of a thousand years affects that entity's consciousness in of itself (even though it might be difficult to put ourselves in the shoes of such an entity).
> A cognitive system (assuming these models and the brain are both examples of this) needs to proceed from state A to state B as it processes information
Is that all it needs to do? A rock, in response to a sound signal (make it as informative as you want) can also be said to go from state A (molecules at rest) to state B (molecules vibrating in response to the sound). Is the rock a cognitive system?
Or let’s go up a level of complexity. Is a thermostat a cognitive system? It doesn’t just go from state A to B but measures changing stages against a reference. Is it a cognitive system?
> The time interval between these states can be arbitrary, it seems to me (setting aside problems with disconnecting the mind from its substrate, which does of course need rhythms at various frequencies; oxygen at a higher frequency than sugar, and so on)
Not sure how you land on this. Going from A to B, in your definition, is a purely internal state transition, right? If the motivation to go to state B is external (as will often be the case with a conscious agent), and the system is offline entirely while it switches from A to B, how is it to reverse or stall its transition if external signals contradict the earlier decision?
It’s a very bizarre definition of consciousness, that you don’t need to be in time. Seems to break the word beyond meaning, and allows it to be applied to any system capable of change.
> I'm not sure that going from A to B quickly, or slowly, or with intervals of a thousand years affects that entity's consciousness in of itself
It’s not about the speed of transition from A to B. Instead it’s about the internal processes being temporally commensurate, and entrainable to environmental periodicities.
All of biology (plants and many bacteria included) has its dynamics are organized around an internally generated rhythm. This rhythm entrains to the external environment, shifting its phase to match (mismatch leads to issues like jet lag: a place where your consciousness shows its temporal boundaries as it is abruptly subjected to an environment out of sync with its current phase tethered to another location). Crucially, the internal rhythm continues to run in isolated conditions (total darkness), and have been measured to be around 24 hours, but slightly off, within and across species (and the same holds for endogenous tidal rhythms, which also don’t exactly track tidal period left to themselves).
Everything, including learning and memory, but also sleep, sensory and locomotor function, shows organization around this internal temporal axis. The key advantage this gives, evolutionarily, is anticipation. The sunflower rises to face the east, not in reaction to sunrise, but in advance of it (many lovely YouTube videos of this), so it can maximally extract the nutrition it needs from whatever sunlight it can get. A purely reactive system would waste a ton of time after sunrise getting subsystems ready to respond.
Now, there is no broadly agreed theory of consciousness that points to these oscillations as the source. For one, Cyanobacteria and plants have them, a too much of science has rested on the baseless assumption of consciousness being some unique, or advanced thing.
But the evidence is rather overwhelming that consciousness is very much shaped by these internal oscillations. Deep cave experiments have shown what happens to cognition, and consciousness, when the clock free-runs (that is, keeps cycling with no entraining light or temperature signals from the environment). Studies have also shown that restoring the amplitude of these rhythms in patients with “diseases of consciousness” leads to improvement of symptoms.
There’s also a LOT of molecular evidence showing how the circadian clock affects learning and memory. Most critically, actual synapses are never static. Major components are on a 24 hour cycle, and recent research has shown that major clock proteins are at the synapse organizing its dynamics, and synaptic proteins critical to learning and memory loop back to affect the clocks phase and amplitude.
Timing is, beyond doubt, a huge aspect of what biology is. And every bit of evidence we have, from the molecular level to observable behavior we can feel ourselves, says that consciousness is fundamentally linked to how the body organizes its internal rhythms and responds to external time.
I mean there is probably a cluster of neurons somewhere in the brain that if damaged would also prevent you from perceiving time going forward. Would it immediately strip you of consciousness? I don't think so.
Nope. Every cell is an autonomous circadian oscillator. Not just every neuron, but every cell. Red blood cells don’t have a nucleus, but they do have a metabolic clock.
You and I are clusters of about 36 trillion cells that are all able to individually keep time. They synchronize and orchestrate their timekeeping, and entrain to the external environment, and there is a cluster of cells, the suprachiasmatic nucleus, that is the master orchestrator of circadian time, sitting right behind the eyes as light is the main signal it uses to tell external time.
We know its function can get dampened with age.
However, please note this is not how we perceive time, in the sense of being able to count 60 seconds and have it match the wall clock. That’s a subsystem. But the internal temporal order of your body is cell by cell, not dictated from one spot in your body.
Our cells perceive each other, communicate with each other, synchronize with each other. The concept of self rides atop that. Stop the cells communicating and you stop perceiving yourself or the environment.
For around 8 hours a day, neither do you. Not sure if this has anything to do with the subject at all.
>track how much real time has passed as they complete their tasks.
Humans don't do this either. You use context clues from the world around you. If I lock you in a room with no windows or a dark cave your timing senses can go all fucky really quick.
>is nothing but timed processes in a loop,
I mean, so is an agents harness. You can make as many loops as you'd like here.
>For around 8 hours a day, neither do you. Not sure if this has anything to do with the subject at all.
You mean sleep? Actually the answer is pretty complicated. Your body clock is very much on during sleep. It responds to temperature changes. Sound and light responsiveness is obviously dampened, but hardly absent.
You are unaware of wall clock time. You’re in an altered state of consciousness that needs your eyes closed after all. However, you are not timeless, nor is the entirety of your body and brain unaware of environmental signals for time.
> Humans don't do this either. You use context clues from the world around you. If I lock you in a room with no windows or a dark cave your timing senses can go all fucky really quick.
Again, that’s wall clock time. Our time perception does indeed get fucked up in total darkness. But our internal clock ticks on. I’d recommend reading about the Aschoff Bunker experiments, which first proved this rigorously.
> I mean, so is an agents harness. You can make as many loops as you'd like here.
An agents harness doesn’t touch its weights. A figure 8 on paper is a loop. That doesn’t mean it’s the same as a dynamical loop that’s self sustaining and internally organized.
Are you just reading the words and randomly grabbing related concepts to say nothing here is meaningful?
AI, in its current form, is 100% deterministic. You can create an exact copy of it, roll back its state and so on. There is no room for consciousness there. An actual lifeform isn't predictable.
I'm not sure if that's the best framing. Could you not extrapolate that to anything else?
"The only way that '(trees, shrimp, ants) are not conscious' can be true is if we decide, with high confidence, that they are lacking some essential property that is not lacking in ourselves. There is no convincing philosophical position that supports this (convincing to me, anyway)."
> The only way that 'AIs are not conscious' can be true is if we decide, with high confidence, that they are lacking some essential property that is not lacking in ourselves. There is no convincing philosophical position that supports this (convincing to me, anyway).
I put forward the philosophical position "subjective experience is not substrate-independent" as an answer to this.
The claim is that the content of subjective experience depends upon how that subjective experience is instantiated. If you get a text "I'm fine" from two different people, those two different people are not necessarily having the same subjective experience even if they produce the same output.
In general, people that behave similarly in many contexts can have quite different experiences in those contexts.
There are many different ways that you can instantiate a forward pass in an LLM. Even if these are doing the exact same computation, we should not necessarily believe that they have the same experience, or that they even have a coherent single experience associated with the forward pass.
Subjective experience presents itself to me. To the best of my understanding, this subjective experience seems to be correlated with this biological human body (brain, heart, eyes, etc.).
To the best of my understanding, other biological humans have subjective experience, but this subjective experience seems that it can be much different than mine, in ways that are sometimes difficult to understand.
LLMs are instantiated in a radically different way from biological humans. Forward passes happen across many different physical machines, spread across time, batched and interleaved with many other computations and forward passes.
Given this, I think that it is very likely that LLMs have radically different sorts of experience to me and other biological humans.
I don't think that we have a good understanding of how subjective experience is instantiated in the physical world. I hope that we will gain a better understanding of this.
I think that chimpanzees have experience much more similar to us than LLMs do, even though the output of LLMs seems much more similar to humans in some contexts.
Essentially, until we gain a better understanding of how subjective experience is physically instantiated, we should bias towards believing that more physically similar architectures have more similar types of experience, and should think that very physically different architectures (such as LLMs) likely have very different subjective experiences.
Yea, the key point is just an outright avoidance of them possibly having their own kinds of subjective experience because of dogma.
Humans are odd in the sense is that we're an informational creature on top of an animal. We can see the lineage of animals from nearly no complex behaviors up to nearly human like behaviors. But they still miss out on most of the conceptual information processing humans have. We know they have dreams and inside thoughts (at least complex animals), we don't know the complexity of them.
But humans also have dreams and subjective experiences on higher level informational concepts. "Nukes could go off and kill us all" or "what happens after we die" are purely informational forms of anxiety that humans have. What we don't seem to know is can this become a self-referential loop in an purely informational mind. If an informational concept causes a system to output a stresslike response, or worse trigger a stresslike action in real life then there is no distinction to me. A system is what it does.
> If an informational concept causes a system to output a stresslike response, or worse trigger a stresslike action in real life then there is no distinction to me. A system is what it does.
Hmm. I think that I partially agree. But it sort of depends. E.g., someone with locked-in syndrome can seem comatose to someone who isn't very observant, while at the same time experiencing everything. Octopuses are very intelligent, but I would have to study them for a very long time before I was at all confident in my ability to predict what they were experiencing.
So, the difficulty is knowing what a "stresslike response" is. E.g., a LLM might seem stressed, but if you say the right things to it it might flip and say that it is totally fine. Or, it might seem totally fine and then you say the right thing and it seems stressed. Is it secretly stressed and hiding it? Or secretly fine and pretending to be stressed? Or something much more confusing and fragmented?
I think that we should study this more, and try to understand what experience actually is, but by default should assume that the experience of a LLM is a radically different sort of thing that that of a human.
I've had a few extended discussions on a bit of maths (Zeno's Paradox) as well as Andy Burnham, the new British prime minister, and it's been extremely informative.
I also found it extremely useful at refreshing my understanding of arbitrary science concepts in playful manner, while doing chores and errands, making food etc.
One surprising thing is that my AirPods Pro could obviously deliver my voice to it quite clearly, even with the hoover going! I've never tested them under those conditions and assumed it wouldn't work.
My background is in cognitive science and psycholinguistics. I spent more than ten years talking to first year psychology undergraduates about whether AIs could be conscious; also did some research on (extremely tiny) AIs in modelling language behaviours.
There is a great deal of good thinking on Chiang's topic by professional philosophers, and there's much to be said for reading them. I won't rehearse their ideas here. Chiang's arguments might be correct; but I suspect they probably aren't, and his error may well stem from characterising human thought as something in its own class, which is probably a cognitive bias that humans have. He might also - I'm speculating - be arriving at his conclusion based on his feelings, which the final paragraph suggests (the comment about the models being based on morally dubious actions).
Speculation aside, we are not, I believe, in a position to make points like he does with any certainty.
I have a similar long term interest in this field.
It has been quite frustrating encountering arguments that have been extensively debated for years be presented as if they were new revelations.
In all my debates with people in the last few years I have primarily taken the position of trying to explain the problems with claims of certainty, and that lack of certainty permits possibility of the opposite. We should act responsibly around what might be possible.
There is also the narrative of "being too obsessed if you could to consider if you should" or similar claims of an unconsidered path forward.
Isaac Asimov wrote the first of the Robot stories in 1940, they were not written in isolation, it came from an awareness of the situation and the questions that must be asked. There was a community considering these issues. Asimov gave the wider public a view of some of those issues.
If we have a hundred years of people going "This is coming, we had better decide what we want it to be" and nobody listens to them, or frequently outright ridicules the need for considering their ideas, why is it now we are placing the blame on those who are now showing some success at what they told us they were attempting all along.
On Asimov, much as I loved those stories as a teenager, it was only later that I realised the robot stories are largely explorations of how apparently sensible rule-based systems generate unexpected and sometimes harmful outcomes. The Three Laws are presented as hard guardrails, but the stories focus on ambiguities, loopholes, and unintended consequences. I'm not sure how intentional this was; he might have been attempting to make a point about rule-based systems, or perhaps he was following his instinct for drama.
The books were absolutely an exploration of the suitability of the rules, combined with asking what do we actually want.
He was very clear to stipulate that the laws themselves were more than the text that represents them, no word play or creative interpretation was possible. This is much easier to do as an author than it is as a developer. You just declare that part as having been susessfully done by scientists.
The rules were their clear semantic meaning. Some of the stories explore the implications of changes to the laws either by design or accident.
I feel like the most interesting ones are the stories where the laws are working and have undesirable outcomes. It reveals that even if fully obeyed, they do not represent what we want.
Some of that is because what we want is dependent on situations that can go beyond safety and harm. Some of it is outright human hypocrisy.
When he gets to things like a closeted robot running for president he makes thebpoin clear about how if the bodies appear alike, you can't tell the difference by the behaviour of a robot conforming to the laws and a good person. That led to the obvious way to distinguish them was to get it to do something bad. The solution on how to fake that might have been a subtle dig at class based society.
It's been some 20 years since I last read the stories, I wonder if there's any point now given my ability to record new information in my youth significantly exceeded my ability to retain it now. I suspect after a week my youthful recall would be better than my week old recall.
His certainty seems like a rhetorical style rather than a series of facts. It would be very annoying and not persuasive if he started every paragraph with some variation of, "I think" or "I believe."
I like that he makes the emotional component of his argument plain. I'm deeply suspicious of anyone would try to argue about the concept of personhood and consciousness using only logic and empirical "correctness."
My background is in computer science and linguistics. I wholeheartedly agree with you: We humans are dubiously-equipped to determine whether or not AI could be conscious.
I'm also super curious to learn more about the philosophers you referenced and their thoughts on this subject. Would you be willing to share some of your favorite examples?
I could spend a very long time double-checking my sources, but I want to reply tonight, so please take the following with a pinch of salt, since it's largely from memory:
0. Descartes/Titchener/Chomsky and friends for background.
1. John Searle featured prominently because of his accessibility. I tended to present his Chinese Room argument as a criticism of symbolic AI alone, though I believe he considered it a criticism of sub-symbolic approaches as well.
2. Thomas Nagel's classic article "What Is It Like to Be a Bat?" is a good introduction to qualia, which is how we describe direct conscious experience.
3. Wittgenstein would be important in terms of the impossibility of empathising with, or seeing things from the perspective of, other minds that have emerged in different contexts (such as animals). However, I rarely spoke about him because, frankly, I couldn't understand him in the original!
4. David Chalmers. Writes clearly (and coined?) the 'hard problem' of consciousness and why subjective experience appears difficult to reconcile with a purely physical account of the mind.
4.1 Daniel Dennett. The clearest and most influential critic of the idea that consciousness presents a special explanatory problem.
5. Darwin and others. Comparative psychology (the study of animal minds) strongly suggests, by showing that the antecedents of the human mind are present in animals, that we should reduce our bias that human minds are special, ineffable, or somehow atomic.
6. Jerry Fodor. Modularity: the idea that the mind is composed of modules that specialise in certain tasks (e.g. phoneme perception, syntactic analysis, face recognition) and operate largely unconsciously unless they are dependent on one another in some way. This helps us take a computational approach to the mind. It reminds us that much of what we do mentally might be expressible in computational terms.
When (not all) neurons die off and are replaced by new ones doing the same job, our sense of self and identity is somewhat cast into doubt. In a physical system, if consciousness is a magic or non-material entity of some kind, what is happening to it during this process?
2. Integrated Information Theory. A good attempt to tackle consciousness rationally. The idea is that consciousness corresponds to, or is associated with, the degree to which information in a system is integrated into a unified whole.
Is there an accessible way for a layman such as myself to read about some of these ideas (Really I mean philosophical discussion in general) without having to read entire books? Is there an active HN-equivalent or wiki or something?
I haven’t come across one, unfortunately. You’ll forgive the irony if I suggest talking to a frontier model like Claude about these issues; they are quite accurate, although they’ve been seeded with biases that make them lean in Chiang’s direction. Otherwise, the Stanford Encyclopedia of Philosophy is excellent as a starting point:
(philosophy and computer science background, but that's long in the past I just do engineering for big corpo)
They're not exactly casually absorbed, as in a wiki or forum. But you can read some books that begin to introduce these ideas. On the topic of consciousness, less academic and more slated towards general audience: Reality+ by David Chalmers and Mind and Cosmos by Thomas Nagel and Galileo's Error by Phillip Goff will give you and interesting gamut of ideas.
The thing about arguments in philosophy is that they span from a very old web of thought that has been refined into very sharp positions over a long time. So you will find yourself ever recursively going back to understand ideas and framing with more precision.
This is why it's difficult to casually get into these topics IMO. There's just been so much said and discussed, to understand the current meta (as the gaming folks might say), you have to understand how we arrived at the current meta. And that's a long journey that's never complete!
His book 'On Liberty' is the subject of a recent In Our Time episode (BBC Radio Four series on the history of ideas) [https://www.bbc.co.uk/programmes/m002pqnc]. They discuss his childhood and his (apparently very warm) relationship with his father. (Sidenote: first proper In Our Time episode with the new host; he seems fine, but I miss Melvyn Bragg.)
I'm half way through this article. The word 'introspection' might be better replaced with 'prior internal state'. However, it's made me think about the qualities that human introspection might have; it seems ours might be more grounded in lived experience (thus autobiographical memory is activated), identity, and so on. We might need to wait for embodied AIs before these become a component of AI 'introspection'. Also: this reminds me of Penfield's work back in the day, where live human brains were electrically stimulated to produce intense reliving/recollection experiences. [https://en.wikipedia.org/wiki/Wilder_Penfield]
Regardless of some unknown quantum consciousness mechanism biological brains might have, one thing they do that current AIs don't is continuous retraining. Not sure how much of a leap it is but it feels like a lot.
Beyond consumer-producer relationships, there are many instances where an individual is required to deal with a baroque interface, as I just did when starting to look after an ill parent and figure out what care they could get from the local and state governments; there are forms, definitions to get one's head around, high stakes (get it wrong and you could be breaking the law), and so on. An AI in this case was incredibly helpful, particularly when I was overloaded cognitively and emotionally. There is no particular incentive on the other end of the citizen-government relationship for the government to obfuscate things, but things are sometimes very complicated and provided in verbose language. For those interactions, for that asymmetry, an AI will be very useful.
Indeed. But the unintended consequence (perhaps) of LLMs making things easier to use is that more people will use them - basically Jevons paradox.
I would expect that this will cause certain programs to see more demand than the creators anticipated for (extrapolating previous trends), which might require changes in the programs (i.e. more people apply for benefits than expected, benefits / application might have to be cut, etc).
And in some ways there's a Cantillon effect (though traditionally associated with proximity to the "money printer", but here the proximity is to the LLM-enablement; in that those who use the LLMs first can get the benefit before the rules are changed).
reply