Aubrey de Grey Discussion, 9

I discussed epistemology and cryonics with Aubrey de Grey via email. Click here to find the rest of the discussion. Yellow quotes are from Aubrey de Grey, with permission. Bluegreen is me, red is other.

Thanks. Hm. I’m sincerely trying my very hardest to understand what you’re saying about your own thought processes, but I’m not making much progress.

I understand. It's very hard. Neither DD nor Popper had much success explaining these things in their books. I mean the books are great, but hardly anyone has thoroughly been persuaded by those books that e.g. justificationism is false.

I'm trying to explain better than they did, but that's tough. It's something I've been working on for a long time, but I haven't yet figured out a way to do it dramatically more effectively than DD and Popper. I think correct epistemology is very important, so I keep working at it. But I'm not blaming you or losing patience or anything like that.

At this point I think where I’m getting stuck is that the differences between your and my descriptions of how you make decisions (and of how one ought to make decisions) mainly hinge on the distinction between (a) not having any further criticisms and (b) not choosing to spend further time coming up with further criticisms,

I think there's a misunderstanding here.

I wouldn't draw a distinction there. If you don't know more criticisms, and resolved all the conflicts of ideas you know about, you're done, you resolved things. Whether you could potentially create more criticisms doesn't change that.

The important thing is not to ignore (or act against) any criticisms (or ideas) that you do know about. Either ones you came up with, or someone told you.

If you do know about a conflict between two ideas, don't arbitrarily pick a side. Rationality requires you either resolve the conflict, or proceed in a way that's neutral regarding the unresolved conflict. This is always possible.

Does that summarize one of my big points more clearly?


In other words, when there's a disagreement, either figure out how to resolve it or how to work around it, but don't assume a conclusion while the debate is ongoing. (The relevant ongoing debate typically being the one in your own mind. This isn't a formula to let irrational confused people hold you up indefinitely. But details of how to deal with this aspect are complex and tricky.)



Secondarily it's also important to be open to criticism and new ideas. If the reason you don't know about a criticism is you buried your head in the sand, that's not OK. (This part is pretty uncontroversial as an ideal, though people often don't live up to it very well.)

and I claim that for most interesting questions that is a distinction that is very hard to make, because it’s almost always fairly easy to come up with a new criticism (and I don’t mean a content-free one like “that’s dumb”, I mean a substantive one). Now, you disagree - you say "It's hard to keep up meaningful criticism for long”. That’s absolutely not my experience. In fact I would go further: I think that the way our brains work is that exhaustion or distraction from what we objectively know we’d like to do is a phenomenon that we generally like to put out of our minds, because we wish it weren’t so, so it’s virtually impossible to know whether we have truly exhausted our potential supply of criticisms. I really, really like to know why I think what I think, so I feel I go further down these rabbit-holes than most people, but they’re still rabbit-holes.

I'm mainly concerned with actual criticisms and conflicts of ideas, not potential.

Apart from the issue of willfully not thinking of arguments you couldn't answer, or choosing not to hear them, then it's only the actual ideas you have that matter and need conflict resolution between them now.

I think the only promising-sounding way to resolve this (i.e. to determine how difficult it really is to keep up meaningful criticism - which will very probably entail gaining a better understanding of each other’s threshold of “meaningful”) is for us to work through a concrete example. Naturally I suggest we continue with cryonics.

I disagree with "only". But that's fine, sure.

Though, actually, I don't think cryonics is ideally suited because on cryonics I'm more in the role of critic, and you more in the role of defending against criticism.

But our epistemology disagreement is kind of along the lines of: I have higher standards. So when I'm in the role of critic, this will come off as: my criticism is picky and demands standards you think can't be met.

If we used a different topic where I have a lot of knowledge and positive claims exposed to criticism, it could more easily be you making criticisms as picky as you want – trying to demonstrate such picky criticisms can't be answered – and then me showing how to answer them.

What do you think?

I reply about cryonics below anyway.

Before that, though, I have a new issue with some of what you said in this latest reply. You seem to have created a massive loophole in your approach here:

- the more you use questions like this and temporarily exclude things due to resource limits, the easier it is to reach agreement. if it's different people, it goes to "since we disagree so much, let's go our separate ways".

I can’t for the life of me see how you can seriously view that as an epistemologically acceptable outcome. And yet, I claim that it is indeed necessary to say that in order to reach your claim that resource limitations are not fatal to the epistemologically respectable method you advocate. Agreeing to disagree is no different from saying “that’s dumb”, except insofar as the participants may have gained a better understanding of the issues (negligibly better, in most cases, I claim). This is particularly important because of the non-level-playing field issue - much more often than not, the two participants in a debate will have unequal resource limits, so one of them will need to quit before the other feels ready to quit, so going separate ways ends up as the only option.

I'm unclear on the problem. If people AGREE to leave each other alone, and act accordingly, then they have a mutually agreeable win/win outcome that neither of them has a criticism of. This resolves the conflict between them that they were trying to sort out.

This doesn't resolve the tough problems in the field – but they know that and aren't claiming otherwise. What their agreement resolves is the problems surrounding their immediate decision making about how to deal with each other.

OK, let’s get back to cryonics.

BTW, what is your explanation of why no one has written good explanations of why to sign up for cryonics anywhere? Why have they left it to you to write it, instead of merely link things?

I think what’s been written by Alcor is (in aggregate) a good explanation, and you’ve read it already, so I didn’t suggest you read it.

In aggregate, I think you will agree it contains flaws. I've pointed some out.

So what's needed to save it is some modifications. Some way to have a position similar to it, without the flaws.

But I've been unable to figure out a position like that. And I haven't found Alcor's material to be much help for doing this.


I'm also unclear on what you think the gist of Alcor's case is. What primary claims make up their argument that you think is good? I actually have very little concept of what you think their website says.

Do you think their website presents something like your argument below? That's not what I got from it.

The evidence you refer to is consistent with infinitely many positions, including ones that conclude not to sign up for cryo. Considering it evidence for a specific conclusion, instead of others it's equally consistent with, is some mix of 1) arbitrary 2) using unstated reasons

Why should a fact fully compatible with non-revivability be counted as "evidence for revivability"?

In most scientific fields, and certainly in almost all of biology, the totality of available evidence is consistent with infinitely many positions, including the position that eating grass cures the common cold.

yes

Thus, one doesn’t reject the position that eating grass cures the common cold on the basis of a boolean approach to available evidence - one does so on the basis, as you said, that the quality of explanations for why eating grass cures the common cold (i.e. refutations of the position that eating grasss does not cure the common cold) is inadequate - there are no “meaningful” such explanations.

i disagree and think one should approach the grass-cures-cold with specific criticisms, not vague quality/justification judgments. Examples below.

Let’s have a go. Grass contains huge numbers of phytochemicals that we have identified, and the limitations of breadth and depth of our investigations are such that we can be quite sure it also contains lots that we have not identified. Phytochemicals have many diverse properties, such as antioxidant properties, that are shared with compounds that are known to have therapeutic effects on the common cold. Kids occasionally eat grass, and they occasionally recover faster than average from the common cold, so in order to know whether grass cures the common cold we would need to survey the cases of this to determine whether the two were positively correlated, and no one has done this. I don’t claim that this is a meaningful refutation of the position that eating grass doesn’t cure the common cold, but I do claim that it is a meaningful refutation of the position that it’s not worth doing the experiment to determine whether eating grass cures the common cold. I don’t claim that it’s a persuasive refutation, but the only reason I have for distinguishing between persuasive and meaningful is probabilistic/justificationist: based on my subjective intuition, I think the chances of the experiment coming out on the side that grass indeed cures the common cold are too low to justify the resources needed to do the experiment. What am I missing?

This argument is fine in the sense of being unlike "that's dumb" with no reason given. It's "meaningful". To put it approximately but perhaps communicate effectively: I wasn't trying to exclude anything even 1% as reasonable as this.

But this passage makes several mistakes. Here are some criticisms:

It's suggesting resources be allocated to this. But it doesn't compare the value it thinks can be gained by this change in resource allocation to the value gained from current allocation. So it doesn't really actually argue its case and is vague about what specifically should be done.

It's too much of a "try this, it might work" approach. There are more promising leads. One way (of many) to get more promising leads is to think of a specific mechanism by which something could work which you don't know how to rule out given current evidence and arguments, and then test that.

Another mistake is looking for correlation itself, when the thing we actually care about is causation (we care whether eating grass CAUSES recovery from colds). A good project would try to determine causation. This could maybe involve looking at correlations, but there'd have to be an idea about what to usefully do with the correlation information if found.


Note BTW that all three of these criticisms use fairly general purpose ideas. They're mildly adapted from previous discussions of other topics. For that reason, it doesn't take much work to create them. And as one builds up a greater knowledge of general purpose criticisms, it gets harder to propose any ideas that pass initial criticism using already-known criticism techniques.

Back to cryonics.

Damage that's hard to see to the naked human eye is not "small" in the relevant sense. The argument is a trick where it gets people to accept the damage is small (physical size in irrelevant regular daily life context), and implies the damage is small (brain still works well).

Why use unaided human eye instead of microscope? It's a parochial approach going after the emotional appeal of what people can see at scale they are used to. Rather than note appearances can be deceiving and try to help the reader understand the underlying reality, it tries to exploit the deceptiveness of appearances.

And it doesn't attempt to explore issues like how much damage would have what consequences. But with no concept of what damage has what consequences, even a correct statement of the damage wouldn't get you anywhere in terms of understanding the consequences. (And it's the consequences like having one's mind still revivable, or being dead, that people care about.)

Sure, all agreed - but they are not making that mistake. It’s known that living systems have pretty impressive self-repair machinery, and that it tends to work better to repair physically smaller damage than physically larger damage. Therefore, even though we know perfectly well that damage too physically small to be seen with the naked eye could still be too much for revivability, we know that there is a whole category of damage that would indeed (probably) be too much and is absent,

ok

and that’s meaningful evidence.

Meaningful evidence – meaning what?

This evidence is consistent with many things, so if you want to bring it up you should give an explanation about what it means. It doesn't speak for itself.

Do you mean that of the infinitely many cryo-doesn't-work possibilities, an infinite subset have been ruled out? Yes. Do you mean that this raises the amount of remaining cryo-does-work possibilities relative to the cryo-doesn't-work possibilities? No, infinity doesn't work that way.

Plus, of course Alcor (and more importantly 21CM) have looked at vitrified tissue with microscopes and not seen appreciable damage

What do you mean "appreciable" and where do they provide this information? Aren't fractures appreciable damage?

How does this fit with Brian Wowk's comments, brought up earlier, about lots of damage? Do you think he was mistaken, or is this somehow compatible?

- but how much magnification is enough? If they were basing everything on 100X microscopic images, what would be your procedure for deciding whether or not to complain that they hadn’t looked at the EM level?

I'd ask WHY they didn't use EM level and see if I see something wrong with their answer. There ought to be an explanation, presumably already written down.

I'd hope the answer wasn't "lack of funds even though it's very important". That'd be a plausible but disappointing answer I could imagine getting.

Not using the best microscopes around would strike me as suspicious enough to ask a question about. But in that scenario, I wouldn't be surprised to find they had a reason I have no criticism of, and then I'd drop it. Advanced technology sometimes has drawbacks in some cases, rather than being universally the best option.

I can certainly provide (as Alcor do) positive evidence for how much damage is tolerable - but of course there are ways to refute it, but only if one views one’s refutations as meaningful. For example, we can look at the amount of variabiity in structure of the brain in non-demented elderly, and we can see big differences between people who are equally cognitively healthy - easily big enough to be seen without a microscope.

Damage and non-damage variation are different things. What is this comparison supposed to accomplish?

People have different ideas. It would unsurprising if this has significant physical consequences since ideas have to have physical form. Though we also can see non-microscopic differences in healthy hearts, lungs, skin, etc, so the easily visible brain differences don't necessarily mean more than those other differences.

You could say, ah, but all one is doing there is identifying changes that are not harmful - but that’s circular, in the absence of direct evidence as to whether the damage done by vitrification is harmful.

I'm unclear what you're saying would be circular, or how you'd answer my comments in the section right above. I think I didn't quite get your point here, unless my comments above address it.

To phrase this as a direct criticism, for the context of me being persuaded, the issues have to be clear to me, so things I find unclear won't work.

To succeed in this context, they have to be either modified to be clear to me (which I always try to do myself before objecting), or else there'd have to be auxiliary explanations, either about the specific subject, or about how to read and think better, so that I could then get the point.

Is that a refutation that you would view as meaningful? If so, what’s your re-refutation of it? And if not, why not?

Yes, meaningful. I think the bar there is real low. I just wanted to exclude complete non-engagement like a tape recorder could accomplish.

Some answers above. Plus this doesn't address some points I raised previously, but we can set those aside for now.

Continue reading the next part of the discussion.