Communications lead at MIRI. Unless otherwise indicated, my posts and comments here reflect my own views, and not necessarily my employer’s.
Rob Bensinger
Karma: 9,589
Thiel on secrets and indefiniteness
Are there any obvious (or, failing that, plausible) large improvements that could have been made (or could be made) to Próspera’s legal code?
[Wishful Thinking]
Possibly-self-fulfilling-prophecy warning!
Re “nudgers”, compare Alex Tabarrok in Ezra Klein’s recent article:
In all of this, the same issue recurs: What should regulators do when there’s an idea that might work to save a large number of lives and appears to be safe in early testing but there isn’t time to run large studies? “People say things like, ‘You shouldn’t cut corners,’” Tabarrok told me. “But that’s stupid. Of course you should cut corners when you need to get somewhere fast. Ambulances go through red lights!”
One problem is no one, on either side of this debate, really knows what will and won’t destroy public trust. Britain, which has been one of the most flexible in its approach to vaccines, has less vaccine hesitancy than Germany or the United States. But is that because of regulatory decisions, policy decisions, population characteristics, history, political leadership or some other factor? Scientists and politicians are jointly managing public psychology, and they’re just guessing. If a faster, looser F.D.A. would lose public trust, that’s a good reason not to have a faster, looser F.D.A. But that’s a possibility, not a fact.
“My view is this was all psychology which no one really understood, so I just said, ‘Go with the expected value. Do the thing that’ll save the most lives and stick with it,’” Tabarrok said. “That’s a better rule than trying to figure out ‘If I do this, what will someone else do?’”
serious harm to children’s development
Why do you think the lockdowns caused serious harm to children’s development?
Illusionism thinks the illusion-of-phenomenal-consciousness is ‘perception-like’ — it’s more like seeing an optical illusion, and less like just having a stubborn hunch that won’t go away even though there’s no apparent perceptual basis for it.
The view you’re describing is different from illusionism, and is more like the one articulated by Dennett for most of his career. E.g., Dennett’s 1979 “On the absence of phenomenology”:
[...] Since I hold that we have privileged access only to judgments, and since I cannot make sense of any claim to the effect that something to which I do not have privileged access is an element of my immediate conscious experience, I am left defending the view that such judgments exhaust our immediate consciousness, that our individual streams of consciousness consist of nothing but such propositional episodes, or better: that such streams of consciousness, composed exclusively of such propositional episodes, are the reality that inspires the variety of misdescriptions that pass for theories of consciousness, both homegrown and academic.
[...] You may be wondering if you even have judgments. Typically these episodes are the momentary, wordless thinkings or convictions (sometimes misleadingly called conscious or episodic beliefs) that are often supposed to be the executive bridges leading to our public, worded introspective reports from our perusal or enjoyment of the phenomenological manifold our reports are about. My view, put bluntly, is that there is no phenomenological manifold in any such relation to our reports. There are the public reports we issue, and then there are episodes of our propositional awareness, our judgments, and then there is—so far as introspection is concerned—darkness. What lies beyond or on the interior of our judgments of the moment, what grounds or causes or controls them, is wholly a matter for science or speculation—in any event it is not a matter to which we have any privileged access at all.
Or his 1991 Consciousness Explained:
[...] You seem to think there’s a difference between thinking (judging, deciding, being of the heartfelt opinion that) something seems pink to you and something really seeming pink to you. But there is no difference.
There is no such phenomenon as really seeming – over and above the phenomenon of judging in one way or another that something is the case.
Indeed, Dennett describes non-physicalism as being based on a “hunch”, as though it were just a nagging hard-to-pin-down belief and not something that feels palpably present in all experience. This seems very weird to me.
These days I believe Dennett endorses illusionism instead, though I’m not sure what changed his mind if so? And I have to wonder whether he has some aphantasia-like condition that made a view as weird as delusionism appealing.
‘You’re allowed to do what you want, provided you don’t learn anything or produce wealth in the process’, vs. ‘You’re allowed to do what you want, provided you learn something or produce wealth in the process’.
No
I think part of the explanation is ‘most consequentialists in professional philosophy dislike utilitarianism’ and ‘there are lots of deontologists too (in general, somewhat more deontologists than consequentialists)’.
I ~agree with this comment. If we do ever want a noun, I’ve proposed error-reductionism. Or maybe we want something more Anglophone… lessening-of-mistake-ism, or something......
General note: I’m confident Luke and Anna wouldn’t endorse Robin’s characterization of their position here. (Nor do I think Robin’s trying to summarize their view in a way they’d endorse. Rather, he’s deliberately reframing their view in other terms to try to encourage original seeing.)
A quick response on Luke and Anna’s behalf (though I’m sure their own response would look very different):
You can call a nuclear weapon design or a computer virus our ‘child’ or ‘descendant’, but this wouldn’t imply that we should have similar policies for nukes or viruses as the ones we have for our actual descendants. There needs to be a specific argument for why we should expect AGI systems to be like descendants on the dimensions that matter.
(Or, if we have the choice of building AGI systems that are more descendant-like versus less descendant-like, there needs to be some argument for why we ought to choose to build highly descendant-like AGI systems. E.g., if we have the option of building sentient AGI vs. nonsentient AGI, is there a reason we should need to choose “sentient”?)
It’s true we shouldn’t mistreat sentient AGI systems any more than we should mistreat humans; but we’re in the position of having to decide what kind of AGI systems to build, with finite resources.
If approach U would produce human-unfriendly AGI and approach F would produce human-friendly AGI, you can object that choosing F over U is “brainwashing” AGI or cruelly preventing U’s existence; but if you instead chose U over F, you could equally object that you’re brainwashing the AGI to be human-unfriendly, or cruelly preventing F’s existence. It’s true that we should expect U to be a lot easier than F, but I deny that this or putting on a blindfold creates a morally relevant asymmetry. In both cases, you’re just making a series of choices that determine what AGI systems look like.
They even seem wary of descendants who are cell-by-cell emulations of prior human brains, “brain-inspired AIs running on human-derived “spaghetti code”, or `opaque’ AI designs …produced by evolutionary algorithms.” Why? Because such descendants “may not have a clear `slot’ in which to specify desirable goals.”
I think Robin is misunderstanding Anna and Luke here; they’re talking about vaguely human-brain-inspired AI, not about literal human brains run on computer hardware. In general, I think Robin’s critique here makes sense as a response to someone saying ‘we should be terrified of how strong and fast-changing ems will be, and potentially be crazily heavy-handed about controlling ems’. I don’t think AGI systems are relevantly analogous, because AGI systems have a value loading problem and ems just don’t.
2012 Robin Hanson comment on “Intelligence Explosion: Evidence and Import”
This is being cute, but I do think parsing ‘effective altruist’ this way makes a bit more sense than tacking on the word ‘aspiring’ and saying ‘aspiring EA’. (Unless you actually are a non-EA who’s aspiring to become one.)
I’m not an ‘aspiring effective altruist’. It’s not that I’m hoping to effectively optimize altruistic goals someday. It’s that I’m already trying to do that, but I’m uncertain about whether I’m succeeding. It’s an ongoing bet, not an aspiration to do something in the future.
‘Aspiring rationalist’ is better, but it feels at least a little bit artificial or faux-modest to me—I’m not aspiring to be a rationalist, I’m aspiring to be rational. I feel like rationalism is weight-training, and rationality is the goal.
If people are unhealthy, we might use ‘health-ism’ to refer to a community or a practice for improving health.
If everyone is already healthy, it seems fine to say they’re healthy but weird to say ‘they’re healthists’. Why is it an ism? Isn’t it just a fact about their physiology?
Yeah, I’m an EA: an Estimated-as-Effective-in-Expectation (in Excess of Endeavors with Equivalent Ends I’ve Evaluated) Agent with an Audaciously Altruistic Agenda.
I’d be interested to hear a steelman of ‘social shame’, which I think is something you’re gesturing at — that, regardless of what Logan’s talking about, a utopian society would make important use of ‘social shame’ to make things go well. (Perhaps that’s not your point, but it’s at least a thing I’m curious about!)
I’d need more evidence/arguments in order to be persuaded of any of these claims:
History, there’s been no such thing as private ‘shame’.
Private ‘shame’ is synonymous with ‘guilt’. (I think these are close together in concept space, but I strongly guess you’re missing important shades of meaning if you’re rounding off Logan’s topic to ‘guilt’. See also In Defense of Shame.)
In general, I think you should ask more initial questions and make fewer assumptions about what’s in Logan’s head. E.g., you’re assuming a very moralistic perspective on Logan’s part (that they’re really talking about guilt, which is really about “when we do something morally bad, or fail to do something morally good”), but Logan is a very aesthetics-oriented, very anti-morality sort of thinker.
Logan prefers the pronoun “they” over “he”, FYI.
This flies directly in the face of the historical record. Until modernity began, shame was explicitly and strategically a public tool almost everywhere.
I don’t think you understood what Logan was saying. Maybe you took “break it” to mean “cause it to have no effect”, where Logan meant something like “cause it to have bad effects that don’t capture the valuable things about shame”.
It’s apparently not true that 90% of startups fail. From Ben Kuhn:
According to an answer on Quora, “the real percentage of venture-backed startups that fail—as defined by companies that provide a 1X return or less to investors—has not risen above 60% since 2001” (2017 source). Other Quora answers I saw claimed numbers as high as 98% (on different definitions, like ten-year “failure” (survival?) rate), but didn’t cite their sources.