Posts
Comments
MCMC is a simpler example to ensure that we’re on the same page on the general topic of how randomness can be involved in algorithms.
Thanks for clarifying :)
Are we 100% on the same page about the role of randomness in MCMC? Is everything I said about MCMC super duper obvious from your perspective?
Yes.
If I run MCMC with a PRNG given random seed 1, it outputs 7.98 ± 0.03. If I use a random seed of 2, then the MCMC spits out a final answer of 8.01 ± 0.03. My question is: does the random seed entering MCMC “have a causal effect on the execution of the algorithm”, in whatever sense you mean by the phrase “have a causal effect on the execution of the algorithm”?
Yes, the seed has a causal effect on the execution of the algorithm by my definition. As was talked about in the comments of the original post, causal closure comes in degrees, and in this case the MCMC algorithm is somewhat causally closed from the seed. An abstract description of the MCMC system that excludes the value of the seed is still a useful abstract description of that system - you can reason about what the algorithm is doing, predict the output within the error bars, etc.
In contrast, the algorithm is not very causally closed to, say, idk, some function f() that is called a bunch of times on each iteration of the MCMC. If we leave f() out of our abstract description of the MCMC system, we don't have a very good description of that system, we can't work out much about what the output would be given an input.
If the 'mental software' I talk about is as causally closed to some biophysics as the MCMC is causally closed to the seed, then my argument in that post is weak. If however it's only as causally closed to biphysics as our program is to f(), then it's not very causally closed, and my argument in that post is stronger.
My MCMC code uses a PRNG that returns random floats between 0 and 1. If I replace that PRNG with return 0.5, i.e. the average of the 0-to-1 interval, then the MCMC now returns a wildly-wrong answer of 942. Is that replacement the kind of thing you have in mind when you say “just take the average of those fluctuations”?
Hmm, yea this is a good counterexample to my limited "just take the average of those fluctuations" claim.
If it's important that my algorithm needs a pseudorandom float between 0 and 1, and I don't have access to the particular PRNG that the algorithm calls, I could replace it with a different PRNG in my abstract description of the MCMC. It won't work exactly the same, but it will still run MCMC and give out a correct answer.
To connect it to the brain stuff: say I have a candidate abstraction of the brain that I hope explains the mind. Say temperatures fluctuate in the brain between 38°C and 39°C. Here are 3 possibilities of how this might effect the abstraction:
- Maybe in the simulation, we can just set the temperature to 38.5°C, and the simulation still correctly predicts the important features of the output. In this case, I consider the abstraction causally closed to the details temperature fluctuations.
- Or maybe temperature is an important source of randomness for the mind algorithm. In the simulation, we need to set the temperature to 38+x°C where, in the simulation, I just generate x as a PRN between 0 and 1. In this case, I still consider the abstraction causally closed to the details of the temperature fluctuations.
- Or maybe even doing the 38+x°C replacement makes the simulation totally wrong and just not do the functions its meant to do. The mind algorithm doesn't just need randomness, it needs systematic patterns that are encoded in the temperature fluctuations. In this case, to simulate the mind, we need to constantly call a function temp() which simulates the finer details of the currents of heat etc throughout the brain. In this case, in my parlance, I'd say the abstraction is not causally closed to the temperature fluctuations.
I'm not saying anything about MCMC. I'm saying random noise is not what I care about, the MCMC example is not capturing what I'm trying to get at when I talk about causal closure.
I don't disagree with anything you've said in this comment, and I'm quite confused about how we're able to talk past each other to this degree.
The most obvious examples are sensory inputs—vision, sounds, etc. I’m not sure why you don’t mention those.
Obviously algorithms are allowed to have inputs, and I agree that the fact that the brain takes in sensory input (and all other kinds of inputs) is not evidence against practical CF. The way I'm defining causal closure is that the algorithm is allowed to take in some narrow band of inputs (narrow relative to, say, the inputs being the dynamics of all the atoms in the atmosphere around the neurons, or whatever). My bad for not making this more explicit, I've gone back and edited the post to make it clearer.
Computer chips have a clear sense in which they exhibit causal closure (even though they are allowed to take in inputs through narrow channels). There is a useful level of abstraction of the chip: the charges in the transistors. We can fully describe all the computations executed by the chip at that level of abstraction plus inputs, because that level of abstraction is causally closed from lower-level details like the trajectories of individual charges. If it wasn't so, then that level of abstraction would not be helpful for understanding the behavior of the computer -- executions would branch conditional on specific charge trajectories, and it would be a rubbish computer.
random noise enters in
I think this is a big source of the confusion, another case where I haven't been clear enough. I agree that algorithms are allowed to receive random noise. What I am worried about is the case where the signals entering the from smaller length scales are systematic rather than random.
If the information leaking into the abstraction can be safely averaged out (say, we just define a uniform temperature throughout the brain as an input to the algorithm), then we can just consider this a part of the abstraction: a temperature parameter you define as an input or whatever. Such an abstraction might be able to create consciousness on a practical classical computer.
But imagine instead that (for sake of argument) it turned out that high-resolution details of temperature fluctuations throughout the brain had a causal effect on the execution of the algorithm such that the algorithm doesn't do what it's meant to do if you just take the average of those fluctuations. In that case, the algorithm is not fully specified on that level of abstraction, and whatever dynamics are important for phenomenal consciousness might be encoded in the details of temperature fluctuations, not be captured by your abstraction.
If you believe there exists "a map between the third-person properties of a physical system and whether or not it has phenomenal consciousness" you believe you can define consciousness with a computation.
I'm not arguing against the claim that you could "define consciousness with a computation". I am arguing against the claim that "consciousness is computation". These are distinct claims.
So, most people who take the materialist perspective believe the material world comes from a sort of "computational universe", e.g. Tegmark IV.
Massive claim, nothing to back it up.
This person's thinking is very loosey-goosey and someone needed to point it out.
when you define the terms properly (i.e. KL-divergence from the firings that would have happened)
I think I have a sense of what's happening here. You don't consider an argument precise enough unless I define things in more mathematical terms. I've been reading a lot more philosophy recently so I'm a lot more of a wordcell than I used to be. You are only comfortable with grounding everything in maths and computation, which is chill. But my view is that maths and computation are not the only symbols upon which constructive discussion can be built.
If you actually want to know the answer: when you define the terms properly (i.e. KL-divergence from the firings that would have happened), the entire paradox goes away.
I'd be excited to actually see this counterargument. Is it written down anywhere that you can link to?
I don't really understand the point of this thought experiment, because if it wasn't phrased in such a mysterious manner, it wouldn't seem relevant to computational functionalism.
I'm sorry my summary of the thought experiment wasn't precise enough for you. You're welcome to read Chalmers' original paper for more details, which I link to at the top of that section.
I also don't understand a single one of your arguments against computational functionalism
I gave very brief recaps of my arguments from the other posts in the sequence here so I can connect those arguments to more general CF (rather than theoretical & practical CF). Sorry if they're too fast. You are welcome to go into the previous posts I link to for more details.
and that's because I think you don't understand them either.
What am I supposed to do with this? The one effect this has is to piss me off and make me less interested in engaging with anything you've said.
You can't just claim that consciousness is "real"
This is an assumption I state at the top of this very article.
and computation is not
I don't "just claim" this, this is what I argue in the theoretical CF post I link to.
You haven't even defined what "real" is.
I define this when I state my "realism about phenomenal consciousness" assumption, to the precision I judge is necessary for this discussion.
most people actually take the opposite approch: computation is the most "real" thing out there, and the universe—and any consciouses therein—arise from it
Big claims. Nothing to back it up. Not sure why you expect me to update on this.
how is computation being fuzzy even related to this question? Consciousness can be the same way.
This is all covered in the theoretical CF post I link to.
Could you recommend any good (up-to-date) reading defending the neuron doctrine?
How would the alien know when they've found the correct encoding scheme?
I'm not sure I understand this. You're saying the alien could look at the initial conditions, since they're much simpler than the quantum fields as the simulation runs? In that case, how could it track down those initial conditions and interpret them?
Ah I see, thanks for clarifying.
Perhaps I should have also given the alien access to infinite compute. I think the alien still wouldn't be able to determine the correct simulation.
And also infinite X if you hit me with another bottleneck of the alien not having enough X in practice.
The thought experiment is intended to be about in-principle rather than practical.
Hmm I guess I gave your original comment too shallow a reading, apologies for that.
So to be clear, are you saying that, if a half-awake version of you looks at a button saying "I am conscious", thinks to themselves "am I conscious? Yes I am!", and presses the button, whether or not that half-awake version was actually correct with that introspection is up to interpretation? In other words, you don't trust the report of your half-awake self?
My instinct is to say something like: if your half-awake self is actually capable of introspecting on their experience in the same way as your awake self is, and has enough executive function to translate that decision into whether or not to press the button, then I would trust their report. But yea, this does tug on my intuition a little bit for sure.
it does not boil down to Chalmer's argument.
As far as I can tell, Scott's argument does not argue against the possibility that a waterfall could execute a single forward pass of a chess playing algorithm, if you defined a gerrymandered enough map between the waterfall and logical states.
When he defines the waterfall as a potential oracle, implicit in that is that the oracle will respond correctly to different inputs - counterfactuals.
Viewing the waterfall's potential oracleness as an intrinsic property of that system is to view counterfactual waterfalls as also intrinsic.
as Aaronson makes clear in the paragraph before the one you quote, where he cites the Chalmers argument!
Different arguments aren't always orthogonal. They are often partial refraimings of the same generators. Maybe I was too clumsy when I said his boils down to the chalmers response, what I really meant to say was his argument is vulnerable to the same issues as the chalmers response (counterfactuals are not intrinsic to the waterfall), which is why I don't think it solves the problem.
if we're talking about playing a single game, the entire argument is ridiculous; I can write the entire "algorithm" a kilobyte of specific instructions.
I don't understand what you're trying to say here.
PS. Apologies that the original response comes off as combative
Thanks, I appreciate this :)
Whether aliens can figure out that fact is irrelevant.
To be clear, would you say that you are disagreeing with "Premise 2" above here?
Premise 2: Phenomenal consciousness is a natural kind: There is an objective fact-of-the-matter whether a conscious experience is occurring, and what that experience is. It is not observer-dependent. It is not down to interpretation. It is an intrinsic property of a system. It is the territory rather than a map.
Is this the passage you're referring to that means I'm "fundamentally misunderstanding computation"?
suppose we actually wanted to use a waterfall to help us calculate chess moves. [...] I conjecture that, given any chess-playing algorithm A that accesses a “waterfall oracle” W, there is an equally-good chess-playing algorithm A0, with similar time and space requirements, that does not access W. If this conjecture holds, then it gives us a perfectly observer-independent way to formalize our intuition that the “semantics” of waterfalls have nothing to do with chess.
This boils down to the Chalmers response. He isn't arguing that the waterfall couldn't implement a single pass through of a chess game, but it couldn't robustly play many different chess games. I discuss the Chalmers response in the appendix and why I think it doesn't fix the issue.
I'm not talking about access consciousness here. I'm not talking about the ability to report. I'm talking about phenomenal consciousness.
Maybe I'm wrong, but I predict you're going to say "there's no difference", or "there's nothing to consciousness besides reporting" or something, which is a position I have sympathy for and is closely related to the I talk about at the end of the post. But reporting is not what I'm talking about here.
Perhaps the same calculation could simulate different real world phenomena, but it doesn't follow that the subjective experiences are different in each case.
I see what you mean I think - I suppose if you're into multiple realizability perhaps the set of all physical processes that the alien settles on all implement the same experience. But this just depends on how broad this set is. If it contains two brains, one thinking about the roman empire and one eating a sandwich, we're stuck.
This also does not follow. Both experiences could happen in the same brain. You - being experience A - may not be aware of experience B - but that does not mean that experience B does not exist.
Yea I did consider this as a counterpoint. I don't have a good answer to this, besides it being unintuitive and violating occam's razor in some sense.
Yea, you might be hitting on at least a big generator of our disagreement. Well spotted
fixed, thanks. Careless exponent juggling
Thanks for the comment Steven.
Your alternative wording of practical CF is indeed basically what I'm arguing against (although, we could interpret different degrees of the simulation having the "exact" same experience, and I think the arguments here don't only argue against the strongest versions but also weaker versions, depending on how strong those arguments are).
I'll explain a bit more why I think practical CF is relevant to CF more generally.
Firstly, functionalist commonly say things like
Computational functionalism: the mind is the software of the brain. (Piccinini)
Which, when I take at face value, is saying that there is actually a program being implemented by the brain that is meaningful to point to (i.e. it's not just a program in the sense that any physical process could be a program if you simulate it (assuming digital physics etc)). That program lives on a level of abstraction above biophysics.
Secondly, computational functionalism, taken at face value again, says that all details of the conscious experience should be encoded in the program that creates it. If this isn't true, then you can't say that conscious experience is that program because the experience has properties that the program does not.
Putnam advances an opposing functionalist view, on which mental states are functional states. (SEP)
He proposes that mental activity implements a probabilistic automaton and that particular mental states are machine states of the automaton’s central processor. (SEP)
the mind is constituted by the programs stored and executed by the brain (Piccinini)
I can accept the charge that this still is a stronger version of CF that a number of functionalists subscribe to. Which is fine! My plan was to address quite narrow claims at the start of the sequence and move onto broader claims later on.
I'd be curious to hear which of the above steps you think miss the mark on capturing common CF views.
thanks, corrected
If I understand your point correctly, that's what I try to establish here
the speed of propagation of ATP molecules (for example) is sensitive to a web of more physical factors like electromagnetic fields, ion channels, thermal fluctuations, etc. If we ignore all these contingencies, we lose causal closure again. If we include them, our mental software becomes even more complicated.
i.e., the cost becomes high because you need to keep including more and more elements of the dynamics.
The statement I'm arguing against is:
Practical CF: A simulation of a human brain on a classical computer, capturing the dynamics of the brain on some coarse-grained level of abstraction, that can run on a computer small and light enough to fit on the surface of Earth, with the simulation running at the same speed as base reality, would cause the conscious experience of that brain.
i.e., the same conscious experience as that brain. I titled this "is the mind a program" rather than "can the mind be approximated by a program".
Whether or not a simulation can have consciousness at all is a broader discussion I'm saving for later in the sequence, and is relevant to a weaker version of CF.
I'll edit to make this more clear.
Yes, perfect causal closure is technically impossible, so it comes in degrees. My argument is that the degree of causal closure of possible abstractions in the brain is less than one might naively expect.
Are there any measures of approximate simulation that you think are useful here?
I am yet to read this but I expect it will be very relevant! https://arxiv.org/abs/2402.09090
Especially if it's something as non-committal as "this mechanism could maybe matter". Does that really invalidate the neuron doctrine?
I agree each of the "mechanisms that maybe matter" are tenuous by themselves, the argument I'm trying to make here is hits-based. There are so many mechanisms that maybe matter, the chances of one of them mattering in a relevant way is quite high.
Thanks for the feedback Garrett.
This was intended to be more of a technical report than a blog post, meaning I wanted to keep the discussion reasonably rigorous/thorough. Which always comes with the downside of it being a slog to read, so apologies for that!
I'll write a shortened version if I find the time!
Thanks James!
One failure mode is that the modification makes the model very dumb in all instances.
Yea, good point. Perhaps an extra condition we'd need to include is that the "difficulty of meta-level questions" should be the same before and after the modification - e.g. - the distribution over stuff it's good at and stuff its bad at should be just as complex (not just good at everything or bad at everything) before and after
Thanks Felix!
This is indeed a cool and surprising result. I think it strengthens the introspection interpretation, but without a requirement to make a judgement of the reliability of some internal signal (right?), it doesn't directly address the question of whether there is a discriminator in there.
Interesting question! I'm afraid I didn't probe the cruxes of those who don't expect hard takeoff. But my guess is that you're right - no hard takeoff ~= the most transformative effects happen before recursive self-improvement
Yea, I think you're hitting on a weird duality between setting and erasing here. I think I agree that setting is more fundamental than erasing. I suppose when talking about energy expenditure of computation, each set bit must be erased in the long run, so they're interchangeable in that sense.
Sorry for the delay. As both you and TheMcDouglas have mentioned; yea, this relies on $H(C|X) = 0$. The way I've worded it above is somewhere between misleading and wrong, have modified. Thanks for pointing this out!
Fixed, thanks!
Thanks for the comment, this is indeed an important component! I've added a couple of sentences pointing in this direction.
fixed, thanks!