Post

"We're all worried," as what it means to do research (in my field, Theoretical CS) seems to be shifting, and shifting fast. What to do? Senior researchers must lead by example, knowing that not everything will pan out. What I'm suggesting below may not work everywhere, but here's my own advice: 1/

What does LPU stand for here? 'Lowest publishable unit' or something like that? In physics I heard of this referred to as a 'publon'; one quanta of publishable material.

"Least Publishable Unit" bsky.app/profile/taop...

- go for exposition. Write, write well (if necessary, learn how to). Papers are meant to disseminate knowledge to others. Reading your work, things should "click": this is important, this is novel, this is the key takeaway. We'll no longer be competing on pages of derivations, that "flex" is gone 3/

What if the AIs write well?

(Technical skills will still be necessary of course, just no longer sufficient) - aim for the hard or interesting stuff (especially conceptual). Now is the time to go for the long, blue-sky-ish projects. Learn new areas, go deep. The times of competing on superficial breadth are gone (sorry, me) 4/

- Relatedly, go for something you are deeply passionate or curious about: if the next couple years are going to be (possibly) a struggle, at least don't let it be about things you don't really like or care about! 5/ (end of personal, biased advice for junior TCS researchers)