Blimey - three conferences in a row is a bit of a strain. Last week, it was Norms and Analysis then Probability, now it's the AAP. Here are a few of my highlights so far.
Roy Sorensen gave a talk on whether you should open an envelope that you know to contain misleading evidence which will dislodge a piece of knowledge that you now possess. This is a very interesting question - or rather, there is a swathe of interesting questions in the vicinity, and I'm not quite sure I managed to pin Roy on which one he intended to ask. One is about the epistemic rationality of opening the envelope, but it is controversial to suppose that norms of epistemic rationality apply to actions like opening envelopes (as opposed to beliefs and degrees of belief). Another batch concerns various instrumental norms: what you should do if you want to maximize your knowledge, what you should do if you want to make sure you're taking account of all the evidence, etc. But the answers to these are kind of obvious. Another one conerns practical rationality, but Sorensen told us he intended a different question to this (at least initially).
Al Hajek gave a fun (and very informative, for me anyway) paper on relationships between the debates about, on the one hand, the claim that the probability of a conditional is the corresponding conditional probability and, on the other, the claim that the expected value of A is the probability of 'A is good'.
David Braddon-Mitchell and Caroline West discussed their - courageous! - view that personal identity over time is a matter of caring about one's future stages. So drastically failing to care about your future stages means you do not continue to exist. Total imprudence is impossible - conceptually impossible by their lights, in fact. One thing I didn't get clear on (I should remember to ask them about this) is whether merely caring that one have some future stages is supposed to count, or whether I must have desires about them in some more robust sense.
Reports from the AAP soon ... and hopefully some photos!