In consideration of the (still questionable) possibility of a true "AI revolution," in which most white-collar work is replaced by AI, it seems to me that such a revolution would be no different from any other economic paradigm shift. These always involve periods of instability as we respond to the new balance but ultimately always produce greater wealth. The main thing that can screw this process up is government intervention, both in the form of existing policy and new policy created to address the crisis. For example, we could conceive of a not-too-distant future in which everyone has access to the legal resources currently available only to corporations, governments and the super-rich. Except that this would almost certainly run afoul of restrictions as to who can practice law.

Well, yeah. The lyrics are awesome, no complaints there :-)

TIL about BTC Infinity Day. Also, happy b-day!

I own Pros & Cons (on vinyl, no less!) but I'll confess I'm not wild about it.

Resolving my own question on dealing with divisive subjects: not a problem, everyone here is a crypto-anarchist ;-)

Holding out for a personal back-yard reactor.

I'm back to trying to implement npub signatures (Schnorr signatures, IIUIC) in Crack (https://crack-lang.org). It's very interesting how much npub borrows from the BTC world, and I'm finding myself pondering a lot on some of the design choices. For example, a number of elements that are folded into a signature are "tagged hashes." These are hashes over both the content and a tag indicating the semantics of the content (for example, a tag of "BIP0340/aux" for the random number used to compute the signature). I can only assume that a lot of this is to ensure that things "break cleanly." So it would be difficult to produce an algorithm that works in test cases but can fail more generally in the wild. It would be interesting to read through the decisions that motivated a lot of this design.