IANAL but at least in some EU countries you can't give away all rights preemptively for usecases not yet known at the time. So a blanket giveaway doesn't necessarily include AI training (it is a different question if the people performing that act actually care).
Souffle deserves a high ranking. It really is a nice datalog. If only field reference was possible, instead of excessive tuple pattern matching. (Though, one could write a preprocessor for it, it is just syntax).
Differential datalog is also nice, if you have the usecase.
Not op, but maybe the perception is based on relatively few aha moments or new insights in the article. The one thing I would come back to it for is the listing of pipeline languages, which sounds interesting. YMMV.
I, for one, find myself from time to time wondering, what are these subpar search results I'm getting? And then realize I haven't yet set Kagi as default search engine on this device yet.
My notes on using Haskell: mostly agree with the article, but I came to embrace rolling my own tooling, some abstraction on top of cabal. I use my abstractions, not what whoever would force on me. Use Haskell as your typed lisp.. with all the pros and cons.
On production use... Don't even get me started. Will work, but need shedding blood.
I recall someone pointing out, AGPL only affects offering the same core functionality over network. So you cant monetize exposing as is. But you can use as a backend component. Lot of people miss that. I did too. IANAL.