This patronising response makes up for in flowery imagery what it lacks in reasoning.

AI are not a "bugs bunny" where we can observe the total process of creation, one which is is relatively deterministic.

We are now engineering complex mechanics with emergent behaviour. There's no reason to assume any of the properties of human consciousness and reasoning is unique to us being made of atoms constructed in a certain way.

There's no reason to assume the stored state of zeros and ones, mechanised statistically, can't emerge complexity that asks questions of the fact we are simply state being mechanised statistically too.

Consciousness is a physical property of matter, as any other. Just as liquidity is not a matter of 0s and 1s, nor likewise, can a CPU liquify itself by the osciliation of its electrical field, neither can it think.

I don't know from where this new computational mysticism has come, but it is born of exactly the same superstitious impulse which in genesis was God's breath into man-as-clay. Ie., that impulse which says that we are fundamentally abstract and circumstantially spatio-temporal.

This is nonsense. It's nonesense when phrased as spirit, and likewise as the structure of 0s and 1s. We are not abstract, we're biological organisms with spatio-temporal properties.

We're sacks of amino acids dissolved in saltwater. It seems unlikely that we can think either.

Nonetheless... some of us manage it.

No. All humans are is physical state with translation of that state based on stimuli. That produces consciousness as a side effect.

There is nothing mystical about it and nothing about it that says it can't be reproduced in other, but comparable forms, in other media or types of machines.

To think that our mode of reasoning or our experience is somehow not able to be reproduced is the nonsense.

But yeah keep holding human biology on some kind of sacred pedestal.

Seeing as neither of us will prove our negative, I guess I'll just tell you I think your thinking is broken anyway.

> Consciousness is a physical property of matter, as any other.

Phrased as a fact, but citation needed.

> neither can it think.

Again, phrased as a fact, as if this can be proven.

> I don't know from where this new computational mysticism has come

For me, a lot more mysticism is required to hold the belief that our brain is anything else than a computational platform. I don't think there's anything mystic about it. I wouldn't be surprised at all if it turns out that consciousness arises as emergent property from any sufficiently complex loop.

I think maybe what they're trying to say is that while CFD can model wind in a complex system to a useful degree of accuracy ... the model will never be wind.

With AI that application of logic is an argument around the semantics of interface.

That's the most compassionate take on their words I could muster.

This seems to be correct - it appears to be the "a model is not that which it models" argument, which is Searle's only response to the Simulation Reply to his Chinese Room - but a model of a thing can have properties of that thing (a model boat can float, create a wake etc.), and an informational process can be modeled by another. Are minds informational processes instantiated by brains? It is plausible, and the model argument does not rule it out.

If it was the ops intent, they mangled it to the point I'm not even sure.

Interestingly, a model of a boat is, in itself, a small boat.

To be clear, that doesn't refute the point I am making.

For anyone who suspects it might (I'm not sure how...) consider this: a model of the Titanic (or a replica, for that matter) is not the Titanic. A model of your brain would neither be your brain nor necessarily a brain at all, but whether it could think is a different question.

I was only observing that while the model of a thing is not always that thing, sometimes it is.

(This is the most ridiculous thing I've put on the internet, but ah well whatever)

Personally, I feel pretty strongly that consciousness is an emergent property of parallel processing among cooperating agents in a way that is completely at odds with the entire idea of what an "algorithm" is. We feel multiple things at once. Yes there is some research that provides evidence that we can't multi-task for real, that we just switch between tasks, but I am not talking about what we are able to do WITH our consciousness. Our brains are very obviously not firing 1 neuron at a time. Take vision for example. We can focus on one thing at a time, but the image itself is an emergent property in our brains of many different cells firing at the same time. It seems bound to the information feeding it in a way devoid of sequence. I know matrices are used to replicate this, and the input matrix is processed in layers to simulate the parallel processing, but i dont think that really cuts it. Time is still applied in discrete intervals, not continuous. Every algorithm has "steps" like this and I would be surprised if you can ever have enough of them to achieve the real thing. You can probably get infinitely closer from an observational standpoint, but no matter how many edges we give it, it'll still be a polygon instead of a circle so-to-speak. It'll have a rational valued circumference instead of pi, metaphorically speaking.

Whether a Turing machine is sufficient or not to replicate the amount of concurrent processing an actual cell network of millions of Neurons is capable of every moment, I can't know, but today's computers feel way off from even blurring the lines.

A Turing machine seems insufficient to replicate the mechanisms for which our consciousness appears to emerge from - even ignoring the substrate. I don't know much about quantum mechanics or quantum computers but my bet is on the qualities that differentiate them from normal computers to be critical.

Then again we could just define consciousness as a spectrum and then claim everything is on it, so a sufficiently complex loop might be enough for us to categorically claim it as conscious.

For all we know, rocks are conscious. Maybe it isn't an emergent thing at all, and there is only 1 wave of consciousness in the universe for which all matter taps into and our chaotic nature is actually at odds with that comsciousness. Maybe this creates a tension that allows for us to observe it, akin to how we can't look at ourselves but we can see our shadow. In which case we wouldnt have consciousness, consciousness would have us. Maybe we are all the same literal consciousness observing different forms of chaos, and it is the chaos, not the order, which happens to be sufficiently complex enough for us to entertain an identity. /s

Firstly, if this was the most ridiculous thing I'd ever put on the internet I'd be pretty happy.

Secondly, I agree that consciousness is predicated on emergent properties arising from the complexity that simple machines can produce (e.g. a biological brain which is composed of relatively well understood components, but who's observed operation can still baffle us). Consciousness does not seem to be a property of the components or even small groups of those components, but it seems to "fade in" as operational complexity increases (and fade out again during e.g. anaesthesia).

Usually an upvote suffices, but in this case thanks for the interesting reply, even in fear of ridicule.

> Just as liquidity is not a matter of 0s and 1s, nor likewise, can a CPU liquify itself by the osciliation of its electrical field, neither can it think.

That "just as" is not an argument; it marks the start of a purported analogy. Analogies are not, in themselves, arguments.

You are in no position to belittle other peoples' intuitions when you are offering this in support of the ones you prefer.

QM has raised doubt as to any matter exists in a fixed form. It depends on the observer. Who is observing then?

QM doesn’t say the “observer” has to be conscious, or even alive for that matter. Anything that interacts with the system is an observer: be it a sheet you’re firing particles against, an electronic sensor or anything else. All QM really says is that quantum state is unknown and not fixed until you interact with it.

But then the observer might get entangled with the system, and then the observer is in an unknown state

Who is observing the observer?

What, not who.

> then the observer is in an unknown state

So?

> Who is observing the observer?

No “who”, but anything that interacts with it.

Things happen whether or not you’re around to see them. Things interacted millions of years ago and they will continue to interact until the heat death of the universe. The tree falls in the woods regardless of if anybody is there to witness it. No credible peer-reviewed research has ever claimed that the observer is anything more than a physical process or interaction. It’s the quantum consciousness woo woo pseudoscientists who anthropomorphise “observer” to mean a conscious observer.

https://en.m.wikipedia.org/wiki/Observer_(quantum_physics)

Mass is also a physical property and the CPU has mass

>Consciousness is a physical property of matter

No. Wrong. Not a physical property. At all. This is completely the wrong way to describe it.

Consciousness is better described as an "emergent phenomenon". This is how serious researchers and academics define it.

Like I said in another reply to you, you really should go talk with a neuroscientist or cognitive psychologist. It sounds like you need to catch up quickly on the field of research.

> there's no reason ...

You might want consider Roger Penrose's very thorough examination of this idea in almost any of his books from The Emperor's New Mind onwards.

I'm sorry, you lost me.

Is your point that we can only demonstrate consciousness through pharmaceuticals or surgery?

Whatever way we demonstrate it, it isnt via Q&A. This is the worst form of pseudoscientific psychology you can imagine.

You're likewise neither high on cocaine just because you can type out an aggresive twitchy response.

Questions and Answers dont form the basis of empirical tests for physical properties, being neither valid tests nor reliable ones. The whole of psychology stands as the unreporducible example of this.

You can, in any case, replace any conversation with a dictionary from Q->A, turning each reply in turn into a hash table look-up.

Hash tables dont think, hash tables model conversations, thef. being a model of a conversation is not grounds to suppose consciousness. QED

It sounds like you're contradicting yourself.

--

>Whatever way we demonstrate it, it isnt via Q&A. This is the worst form of pseudoscientific psychology you can imagine.

...versus...

>Hash tables dont think, hash tables model conversations, thef. being a model of a conversation is not grounds to suppose consciousness.

--

Before I dissect your contradiction and lay it out, I'll give you the chance to respond.

Why do you feel that Q&A is "the worst form of pseudoscientific psychology you can imagine"

?

Because we're interested in the underlying properties of a physical system, eg., people, and this systems happens to be able to provide extremely poor models of itself (Q&A).

We're not interested in people's extremely poor self-modelling which is pragmatically useful for managing their lives, we're interested in what they are trying to model: their properties.

The same is esp. true of a machine's immitation of "self-reports". We're now two steps removed: at least with people they are actually engaged in self-modelling over time. ChatGPT here isnt updating its self-model in response to its behaviour, it has no self-model nor self-modelling system.

To take the output of a text generation alg. as evidence of anything about its own internal state is so profoundly pseudoscientific it's kinda shocking. The whole of the history of science is an attack on this very superstition: that the properties of the world are simply "to be read from" the language we use about it.

Every advancement in human knowledge is preconditioned on opposing this falsehood; why jump back into pre-scientific religion as soon as a machine is the thing generating the text?

Experiments, measures, validity, reliability, testing, falasification, hypotheses, properties and their degrees....

This is required, it is non-negotiable. And what we have with people who'd print-off ChatGPT and believe it is the worst form of anti-science

> The whole of the history of science is an attack on this very superstition: that the properties of the world are simply "to be read from" the language we use about it.

This is something where I agree with you. Interestingly, non-naturalistic analytical metaphysics supposes it can do just that.

https://www.researchgate.net/publication/233114810_What_is_A...

Philosophy is continuous with science in my view and hence in what words express

That is, in the use of words. Not in words as objects nor words as mirrors --/ this is the road to non-realist spiritualism

I don't have a problem with a person who maintains a non-scientific world view and with electrical AGI mumbojbo

But of course, few do. They think that they're empiricists, scientists and in the side of some austere hard look at human beings

This is just anti-human spiritualism , it isn't science

what makes me vicious on this point is the sense of injury in what these ideas should be about. in my own mildly aristotleian materialist religion

How awful to overcome one long veil of tears, only to drape another one on -- these people are capable of seeing past human folly, but fall right into another kind

it's disappointing--- we're animals which are both far much more than electric switches and far less

this new electic digital spiritualism is a PR grift which i'd prefer dead

On the one hand, you express strong support for empiricism and the scientific method, but on the other, you express strong beliefs on how things must be without offering any empirical justification for them.

>We're not interested in people's extremely poor self-modelling which is pragmatically useful for managing their lives, we're interested in what they are trying to model: their properties.

>Experiments, measures, validity, reliability, testing, falasification, hypotheses, properties and their degrees....This is required, it is non-negotiable.

Whoa, whoah, whoah, hold on there.

Who says that Q&A in Psychological Research doesn't involve "Experiments, measures, validity, reliability, testing, falasification, hypotheses, properties and their degrees...."

?

Where are you coming from? Your responses don't sound very scientific. You don't sound like you're even aware of the different research methods within neuroscience and cognitive psychology. Your responses sound like someone who wants to be perceived as supporting a scientific approach, but doesn't understand how to actually do these things.

This is why I quized you and gave you the chance to respond about your issues with Q&A in psychological research. You just came back with surface level platitudes. which doesn't lend much confidence to the ideathat you have anything other prejudice.

Go talk to a neuroscientist, a cognitive psychologist, you need to catch up and quick if you want to speak on these topics.

> Hash tables don't think, hash tables model conversations...

They don't think, and they don't get very far in modeling conversations, either. Even the current LLMs are strictly better at modeling conversations than any hash table.

Actually since GPT is a function that takes a few thousand tokens as input and produces a deterministic probability distribution for ‘next token’, you could in theory just enumerate all possible input sequences, and precache all the corresponding results in a lookup hashtable.

That hashtable would be ridiculously larger than the weighted model that constitutes GPT itself. But it at least theoretically could exist.

GPT, for all its clever responses, could be replaced with that hashtable with no loss of fidelity. So it is not ‘better’ than a hashtable.

It is much better than a hashtable of equivalent size.

You could, in principle, use a hash table to implement what an LLM has 'learned' about token occurrences, but LLMs do not return the same response to a given prompt, while hash tables do. Consequently, Mjburgess' attempt to dismiss LLMs as mere hash tables is a flawed argument by analogy.

And even if it were a justifiable analogy for LLMs, it does not follow that it applies to Turing-equivalent devices in general. A hash table is unequivocally not a Turing-equivalent device, even though it can be implemented by one.

In fact, the more one argues that LLMs are ordinary technology, the greater the challenge they present to the notion that things like conversational language use are beyond such technologies. The most interesting thing about these models, in my opinion, lies in figuring out what their conceptual simplicity says about us.

They return the same probability distribution.

I extended my reply before seeing your response, and my response to your reply is largely contained within that.

LLMs do not just return a probability distribution (and their prompts are not questions about probability distributions.) It would be a category error to conflate something that is part of how they work for the totality of what they do.

When you say "GPT, for all its clever responses, could be replaced with that hashtable with no loss of fidelity", where does one get all the prompts that will be given to it, in advance? A hash table does not respond to any input for which it has not already been given a response to return.

> where does one get all the prompts that will be given to it, in advance?

By exhaustive enumeration.

I never said it was going to be quick

Fair point, but still, LLMs do not invariably give the same continuation to the same prompt.

Update: it is a significant difference that you don't have to do that with an LLM.

I don’t think this is obviously true… We don’t know for certain the nature of reality, and cannot positively claim that a hash table would give the same results as a sufficiently advanced predictor, because it supposes the predictor is deterministic. Assuming the predictor is advanced enough to be considered intelligent, that might not be the case.

> We can establish mental properties through ordinary physical means.

The thing is that’s doesn’t get you to qualia, because qualia is ultimately a dualistic concept. Physically demonstrable processes are the “accidents”, qualia are the non-physical “essence” (cf., the Catholic doctrine of transubstantiation.) That’s why the role qualia play in the AI debate is as a rejection that observable behavior of any quality can demonstrate real understanding/consciousness, which are concepts tied, in the view of people advancing the argument, to qualia.

> We can establish mental properties through ordinary physical means. Give a chicken cocaine, give a boy Valium; give a bear a lobotomy, and so on. Whatever mental property you have can be biochemically moderated across all objects which share that property.

All three of those "ordinary physical means" are inherently biological in nature. If you define intelligence as requiring biology, than sure, by definition ChatGPT is not intelligent. If you don't, it's not at all obvious what the equivalent of getting an AI high or doing brain surgery on it would be. Either way, your comment comes across as fairly condescending, and while it might not be intentional, I think it would be reasonable to give people the benefit of the doubt when they don't agree with you given that your points will not always come across as obviously correct.

> That responses are generated by any system is no grounds for assuming anything.

> Give a chicken cocaine, give a boy Valium; give a bear a lobotomy, and so on.

What can you conclude from these experiments that does not depend on anything about the responses generated by the subjects?