In the age of AI, choosing boring tech seems like a way to improve your outcomes. That's why all examples from big labs are in either TS or Python. Same applies to infra, dbs, and other stuff.

Conjecture: Embrace boring tech, to build new tech.