A cognitive algorithm for "free will."

I haven't thought it through very carefully, but my hunch is that you're not quite getting the crux of the issue.

For one thing, there's a little compatibilism joke I like:

"I don't see why the chess engine is going through all the effort of evaluating 400,000 possible move sequences. It's running a deterministic algorithm! The answer is predetermined!!"

Hahaha. The moral, I think, is that there are two stories:

• "the chess engine moved the knight because that was the result of a deterministic process involving a bunch of interacting transistors on a chip"
• "the chess engine moved the knight because it evaluated 400,000 move sequences and found that moving the knight was best"

These sound like different stories, but they are in fact the same story told at two different levels of abstraction.

So that's one hint: we can get tripped up by failing to appreciate that stories at multiple levels of abstraction can all be "true" simultaneously.

Then another related hint is a thing I wrote here [LW · GW] about the hard problem of consciousness:

All my internal models are simplified entities, containing their essential behavior and properties, but not usually capturing the nuts-and-bolts of how they work in the real world. (In a programming analogy, you could say that we're modeling the [global neuronal workspace's] API & documentation, not its implementation.) Thus, my attention schema does not involve neurons or synapses or GNWs or anything like that, even if, in reality, that's what it's modeling.

So I think it's a slighty wrong starting point to think along the lines of "the processes in our brains to which we have conscious access are a subset of all the processes in our brains". Instead I think we should be saying, "when we introspect, we're not directly seeing any of the processes in our brains, instead the things we perceive are certain states of certain models that abstract over some of the processes in our brains"—just as "a dog barking" is an abstract model that represents some complicated statistical predictions about incoming photons and sound waves and so on.

I don't really have a conclusion here, and I don't mean for this to be a strong disagreement with what you wrote, just maybe a tweak in the framing or something. :-)

The reason why it feels that you have made this decision of your own “free will” is that you cannot detect the overwhelming evidence against it. The mysterious, rapid, unconscious core brain activity that structures your conscious qualitative experience is unavailable to you.

That is, unavailable to my conscious mind.

What you say is similar to Sam Harris's arguments. Both have two principle problems: that the argument depends on an implicit definition of free will, that it is a conscious choice; and an implicit claim about selfhood, that the conscious mind is the self,and the unconscious mind is not.

The compatibilist definition of free will is concerned with whether you are able to act on your desires ... not on whether your desires are conscious or unconscious.

The libertarian definition of free will is concerned with whether your decisions are free of external determimation, or unpredictable in principle... not on whether they are conscious or unconscious. If the unconscious processes that lead to decisions, and about which you know nothing, are indeterministic, then the case for libertarian free will is strengthened, not weakened.

What you do not remember, because it happened too quickly or unconsciously, was the process by which that selection occurred. Or the operations underneath that

You don't observe them, so you don't observe them as deterministic. And in fact, you don't know that they are determinustic , because that would require empirical investigation which neither you nor Yudkowsky have performed. An assumption of determinism has been smuggled in by a choice of language, the use of the word "algorithm".

The issue of conscious control over decision making is interesting and important as well, but it is at best one third of the whole issue.

I have sort of unlearnt how to think of free will in a nondeterministic sense. As such, I tripped over the part where you said there were "arguments against free will." Like, yes of course the sensation of volition is produced by a deterministic, predictable process; how else could it be about the deciding process? Aboutness only exists in causal systems.

A more interesting question may be what the sensation is for? What part of our high-level cognition depends on noticing that we are making a decision?

A better question is, if you write a complex enough decision making algorithm, would there necessarily be a part of it that would naturally map into the free will quale?

A fundamental difficulty in thinking logically about free will is that it involves thinking logically.

Logic, by its very nature, has embedded as its most essential hidden premise a deterministic structure. This makes all reasoning chains, no matter what their subject (including this one), to be deterministic. In other words, a deterministic structure is imposed upon the elements that will be logically analyzed so that they can be logically analyzed.

This leads one, if they ignore this structure is present as the very first link in chain, then proceeds to analyze the entire chain minus this hidden first premise in an attempt to determine what can be abstracted out from it, to incur into an involuntary 'begging the question' and to conclude all elements present in the chain, and all their mutual relations, are strictly deterministic. And, by extensions, that free will doesn't exist in reality, when the most we can actually say is that free will doesn't exist as a deduced link within deterministically structured logical reasoning chains.

Notice that this doesn't preclude free will from being part of deterministically structured logical reasoning chains, it only says where free will cannot be present. It can still be present as an irreducible axiomatic premise, an "assuming free will exists..." used to reach further deductions. But that's it. Any attempt at moving it from the position of an axiom down into the chain proper will invariably fail because the chain itself doesn't admit of it.

I've always seen it as, the sensation of free will comes from the ability to predict one's own actions before they occur, in such a way as those predictions themselves seem to have causal efficacy to make themselves come true. Some predictions have that feeling - that they themselves are what's making the thing being predicted happen (that the part of the brain doing the predicting is outputting that straight to the part handling the behavior, and this hand-off of information is perceived by other regions) - and others don't.

The former always feel like volition, even when they are "long range" (plans for the future, for instance), but particularly when they are "short range" (like moving one's arm, which has a very volition-ish feeling). It's a bit like imperative, or better yet, performative statements ("I now pronounce you man and wife") in language, which directly cause the thing they are describing.

When you feel as if you did not freely will something, it's always either because you did not predict it, or because your prediction did not include the further prediction that it, itself, was what was causing the event, due to some kind of dissociation between the part of the brain doing the predicting, and the part acting. (When I dream at night, for instance, I rarely have any sense of free will, because part of the nature of dreaming is that predictive mechanisms are relaxed and various portions of the brain are heavily dissociated from one another.)