Texas is also dominating renewable energy production

Thank you, likewise!

The two big changes for me are: 1) a large proportion of more senior people from pre-2020 don't attend anymore, so somehow people my gen are now the senior people?; and 2) AI obviously changed a ton in the intervening years, in both the technology and the structure of the research community.

Thanks for keeping up the thread all five days!

We used to work in the same lab! (For 2-3 years or so; RIP MetaMakers Institute.)

Relatedly, Schaeffer claims that the current paradigm in AI of competitions/leaderboards as a driver of progress started in game AI, w/ the computer chess tournament at the 1970 ACM National Conference being the earliest he can identify w/ the key components (open entry, published rankings, etc.).

The only thing keeping it usable imo is that it's still kind of unix under the hood (I use 'fd' for this, installed via homebrew, formulae.brew.sh/formula/fd)

Update: It turned out the triage was easier than expected, because few applicants even attempted to address the opening in their cover letter (it's a joint CS/physics position, so you have to at least gesture towards why you're both).

For me it goes away when I take off my glasses! Or at least isn't noticeable.

I believe it's a combination of dispersion causing a color-dependent offset and something like a 3d glasses effect causing the brain to reinterpret that as 3d

when listing techniques of classical rhetoric, Aristotle forgot to include "juxtapose on a plot"

accompanying blog post

I've heard some eye popping salary numbers for senior infrastructure people they've been recruiting. Not even MLops people, just anybody with large cluster management experience.