I feel like this is almost deliberately missing the point. My median expectation is that AI boosts the economy enormously, so shorting things would be a terrible idea. But non-trivial tail risk is that it kills everyone. So shorting things would be pointless.

That is from Tom Chivers.  It is easy enough to say buy seriously out of the money puts, and be long with the rest of your portfolio.  But few (if any) of the worriers are doing that.  I should note also that many (most? almost all?) of the bad scenarios have intermediate points of great worry and catastrophe where you can cash in on your puts well before everyone dies.  No leverage required, just spend 5k or 10k a year on this, and if you are wrong consider it a mistaken insurance policy.  If you do not know much about finance, the AGI will guide you in this endeavor.

Or some are saying “Markets are bad at pricing long-term idiosyncratic risk.”  If so, all the more reason to spend on those puts.

You will find many, many mistakes in financial theory when people try to rebut the presumption that, given their views, they should in some way or another be short the market.  And do not just tell me which long positions you hold, those are easy bets, as you can be totally wrong and the long positions still will offer normal, risk-adjusted rates of return.  It is your shorts, whether explicit or implicit, that reveal your true soul.  You may recall that Victor Niederhoffer thought an investor should never go net short on an asset.  I am reluctant to use the word “never,” but my own view is not so different.

Is it really so hard to say (and do): “Thank you, Tyler, I just went out and bought those puts!”?

Apparently so.  And perhaps that is because, deep down, your own intuitions realize that, while some significant costs from AI will appear, it really won’t be that bad after all.

Words to live by.