Small nit: if you actually read your citation, it's not Norvig's law, but Proebsting's law[1]. It's a pessimistic spoof of Moore's law, but IMO just reflects ho

Small nit: if you actually read your citation, it's not Norvig's law, but Proebsting's law[1]. It's a pessimistic spoof of Moore's law, but IMO just reflects how much closer to optimal we started from than the first transistors were. Or maybe how much of a head start we've had doing this on paper for centuries. Like, I would frankly be surprised if we ever got more than one doubling out of Proebsting's law. [1]: https://www.gwern.net/docs/cs/algorithm/2001-scott.pdf https://www.gwern.net/docs/cs/algorithm/2001-scott.pdf