Report comment

A person who accepts induction will quickly become convinced of the utility of deduction: it just keeps on working, over and over. A person who accepts deduction can question induction, but the questioning only has the appearance of seriousness. Induction doesn’t prove its conclusions! – the skeptic objects. Well no one ever said it did; it only makes them probable. See Troy’s reply to this thread on probability.

A skeptical argument that convicts our deductive logic of self-refutation would be a great argument. If it worked. But for first-order logic, we have a soundness and completeness theorem. So no such result is in the offing there, at least.

Tu quoque is ordinarily a fallacy because the shared belief or action has not been shown to be implied by reason. If I criticize you for overeating, and you point out that I eat just as much, the fallaciousness of your reply depends on the openness of eating-this-much to rational criticism. But the inference patterns of rational thought, taken as a whole, are not open to rational criticism (though individually each pattern is open to criticism via coherence with the others – and a lot of good work has been done in that area).

Report comment

2) Do the ‘top posts’ overlap much with the posts that get the most traffic? I’m aware of why which posts are indeed considered ‘top posts’ by Scott himself. If one of his articles is the most shared on social media, gets the most positive feedback from regular readers, and friends, Scott trusts the most, and inspire people, I understand why those are considered among the best. However, I suspect some of the posts Scott has written touching upon social justice, or neoreaction, or related social issues, get the most traffic.

3) When I was visiting Berkeley this past summer, I was sharing my excitement over Slate Star Codex, and a half-seriously ventured the idea that I would like this site to have a donation button, and if it did, I would donate. Concerns raised by others were that if his readership paid him to blog all the time, medicine would be deprived of one of the few statistically literate doctors America sorely needs. Additionally, it was mentioned that changing the incentive structure for how Scott blogs might negatively affect the quality of his posts.

Still, I learned something from asking this question. In conversation with Ben Kuhn, and Katja Grace, at the CFAR alumni reunion this last summer, they were musing over how Scott might be one of the best non-fiction writers currently writing, including among published authors. Some essays, and analyses, written by Scott on this blog have really changed how I think for the better. All of this makes me think that in addition to the intellectual stimulation, and entertainment value, this blog provides, Scott also provides great educational value for what I’m assuming are his thousands of readers.

I wish that whatever think tanks, or publications, which are seeking someone to know how to solve a problem are able to find Scott if the answers he can provide are the fit they’re looking for. I doubt it’s feasible given his current hectic work-life balance, but it seems as if there’s more good Scott could do with targeted, formalized writing. I’ve got no clue if there’s any low-hanging fruit here, though.

Report comment

Of course I totally cracked consciousness long ago (yes, I really am that good). In fact folks, I’ve succeeded in making total mince-meat of virtually all the central problems of philosophy and cog-sci in the course of developing my Mathematico-Cognition Reality Theory (MCRT) over the years.

The review article names 5 general categories of theories of consciousness:

1. a global workspace,
2. information integration,
3. an internal self-model,
4. higher-level representations,
5. attention mechanisms.

So which general theory of consciousness is correct according my MCRT ?

*drum roll please*…… the answer is … 4. higher-level representations!

Consciousness is a system of symbolic representation, the purpose of which is to enable communication between the society of agents that makes up the mind. But higher-order representations of what? The short answer is narratives of goals; consciousness is a system which represents goals in the form of narratives.

Other theories of consciousness, while having some merit, all have serious flaws. Let me briefly explain what I believe to be the flaws in the other theories:

(1) global workspace; there are elements of this, but its not the central feature of consciousness – global workspace is more a description of certain activities of the brain correlated with consciousness, rather than an explanation of consciousness per se.

(2) information integration ; this theory is an interesting attempt at defining consciousness in terms of information theory, not a bad try but plain flat wrong, Scott Aaronson did a good rebuttal of this on his blog

(3) Internal self-model ; internal self-models *do* contribute to consciousness, but this is not the central defining feature of consciousness – for instance it’s possible to have awareness of the external world without any self-awareness.

(5) Attention mechanisms; not really, see (1), this is simply a correlate of consciousness , not the explanation for it.

Report comment

On the other points, (I hope obviously) I’m playing up the “genuine affection” thing not because it is all that matters. Clearly there is more to social skills besides hugs, including dealing with adversarial situations. That said, I think the need for real affection gets underplayed in conversations such as this.

What I mean is, reading the first couple posts on the subthread, the posters seemed to paint social skills as strictly adversarial (frex, dealing with shitty HR). And yes, we need those. Not everyone is a friend. But I think that is such a weak half of real social engagement, and that even getting good at adversarial social skills will miss two things: 1) the genuine rewards of warmth and 2) loyal allies who will go to the matt.

Report comment

Regarding genuinely liking people, no doubt the “charming sociopath” type exists. But how common are they? How likely is it that people on this thread could pull that off?

My money’s on about 2 to 5% — the prevalence of sociopathy, adjusted upward some to account for subclinical cases and good actors. I’m almost sure that being a charming sociopath requires being a sociopath.

That said, there are ways to be cool that fall between that and a genuine delight in everyone you meet. On the third hand, though, the rationalist community’s folk idea of social success seems to focus more on the former, and that’s probably a bad thing. I think the reason for it is mythology: MoR‘s Professor Quirrell is fun to read, but he’s not a very good role model, and not just because he’s evil. A lot of people don’t seem to get that.

Report comment

The attitudes in this thread are social suicide. What I am suggesting is to examine those attitudes themselves.

Regarding genuinely liking people, no doubt the “charming sociopath” type exists. But how common are they? How likely is it that people on this thread could pull that off?

Which, of course, I don’t know. But that is not how I’d place my bets.

Again, attitudes and stances. Compare:

“I wish I could pull off being a charming sociopath”

vs.

“I really like people, and I want them to know I like them, and I want to know when they like me, cuz liking and being liked is AWESOME!”

Which do you want to be good at? Which do you want people to sense from you?

Report comment

“Second, people who have good social skills are often pretty good at detecting condescension and contempt. So if you program has, from its inception, the idea that others are dumb and irrational then you’re baking in failure”

There’s some bad epistemology here. Whether or not other people would be offended by a particular claim doesn’t affect whether or not you should believe it.

Anyway, I do think that there is a negative correlation between the kind of epistemic rationality we talk about here and social/streetsmart skill. I wouldn’t consider that condescending, though; there are many tradeoffs in life.

There’s a tradeoff between being able to reach high shelves and not hitting your head on high ceilings! Yes, it’s called height. Pointing this out is not condescending or contemptuous to tall people. Or to short people.

I also like what you say about “genuinely liking people”, I think there’s a grain of truth to it. Though plenty of people who are good with social/streetsmart skills don’t like most people, but I think they do enjoy interacting with them, if only through the pleasure of getting the better of them.

Report comment

Maybe that is indeed the best system; I just want to note that the criterion of ‘correct reasoning’ that assumes you need infinite chains of justification is less epistemically modest than the one that allows unjustified assumptions. The skeptic and the anti-skeptic who assume it’s ‘bad’ to have externally unjustified axioms are both assuming without warrant a criterion of epistemic ‘badness’ that we should properly be skeptical of. In particular, we should be skeptical of such a criterion if it gives equally bad grades to ‘a calculator that outputs 2+2=5′ and ‘a calculator that outputs 2+2=4 but can’t prove it’.

One alternative to an infinite regress would be to affirm all and only the theorems that are derivable in every logic. Or we could affirm all and only the theorems that meet a formal criterion — e.g., consistency. But, again, why should we accept either of these rules-for-acceptability? And it’s very clear that the logic that proves no theorems means that there are no universally derivable theorems; and it’s also clear that the logic proves no inconsistencies. If your skepticism undermines all your reasons for rejecting the logic that proves no theorems, that’s a reduction to absurdity of your skepticism.

Report comment

Report comment