You can call me Witzvo. My determination of whether I’m a “rationalist” is waiting on data to be supplied by your responses. I found HPMOR hilarious and insightful (I was hooked from the first chapter which so beautifully juxtaposed a rationalist child with all-too-realistic adults), and lurked some for a while. I have one previous post which I doubt ever got read. To be critical, my general impression of the discussions here is that they are self-congratulatory, smarter than they are wise, and sometimes obsessed with philosophically meaningful but not terribly urgent debates. However, I do value the criteria by which karma is obtained. And I saw some evidence of responses being actually based on the merits of an argument presented, which is commendable. Also, Eliezer should be commended for sticking his neck out so far and so often.
I was born into a sect of Christianity that is heretical in various ways, but notably in that they believe that God is operating all for the (eventual) good of mankind, and that we will all be saved (e.g. no eternal Hell). I remain agnostic. Talk about non-falsifiability and Occam’s razor all you like, but a Bayesian doesn’t abandon the possibilities to which he assigns prior mass without evidence, and even then the posterior mass generally just drops towards 0, not all the way. Still, my life is basically secular; I don’t think there’s an important observable difference in how I live my life from how an atheist lives, and that’s pretty much the end of the matter for me. Oh, perhaps I have times of weakness, but who doesn’t?
I have formal training in statistics. I am very sympathetic to the Savage / de Finetti schools of subjective Bayesianism, but if I had to name my philosophy of science I’d call it Boxian, after George Box (c.f. http://www.jstor.org/stable/2982063; I highly recommend this paper AND the discussion. Sorry about the pay walls).
I find the Solomonoff/Kolmogorov/AIXI ideas fascinating and inspiring. I aspire to compute for example, (a computationally bounded approximation to) the normal forms of (a finite subsequence of) a countable sequence of de Bruijn lambda terms and go from there. I do not see any lurking existential crisis in doing so.
In fact, maybe I’ve missed something, but I have not yet identified an actionable issue regarding one of the much-discussed existential crises. I do not participate much in the political system of my country or even see how that would help particularly except and unless through actual rational discussion and other action.
I find far more profit in exploring ideas, such as say, Inventing on Principle (http://vimeo.com/36579366), or Incremental Learning in Inductive Programming (http://www.cogsys.wiai.uni-bamberg.de/aaip09/aaip09_submissions/incremental.pdf), either of which I would be happy to discuss.
I am also intellectually lonely.
That’s probably more than enough. Go on and tell me something less wrong.
The author is far too free with the notion of the Bayesian answer. At the level of common practice there is meta-analysis, which is fraught with problems. There’s subjective Bayesianism, which is fine in principle, but in practice has the same limitations: why should that be my prior? what underlying mechanism can explain all these inconsistently measured results and how do I formulate all those complicating possibilities into a likelihood function? Objective priors are a perennial subject of research in statistics which help somewhat in simple parametric problems. Non-parametric priors (e.g. Dirichlet, Gaussian, Levy, … processes) can be made to work in some cases, but aren’t easy to formulate in statistically efficient, computationally efficient, or even sensible (e.g. statistically consistent) ways, in general. AIXI and Solomonoff priors hold out tantalizing theoretical possibilities, but these are not yet practicable.
The best practice of smart scientists and statisticians in my experience is a process of iterative refinement. (This is true both in their own head and in the conduct of “Science.” Call it a computational shortcut, if you will, designed to accomodate our limited human brains. Importantly, it also let’s us stand on the shoulders of the giants who came before.)
One conducts experiments, build models and hypotheses, tests predictions, and conducts new experiments. Bayesian inference is only a consistent way to procede to the truth if the prior contains mass on that truth. Often reality turns out to be more complicated than we had any business imagining it to be before conducting the experiments, and practicable priors would have missed out on it.
The main point here then is that iteration and testing of models happens as a part of the process of understanding an experiment / analyzing data, not just at the level of designing a sequence of experiments.
Box’s 1980 paper contains much wisdom still relevant today to the aspiring rationalist: http://www.cs.princeton.edu/courses/archive/fall11/cos597C/reading/Box1980.pdf