Yeah I'm struggling to recall any other recent instances of companies giving their "AI" (really LLM) ambitions a realistic scope. Hearing "It's useful for some specific things" from Kagi, rather than "it will be in everything and change everything" is ... shockingly refreshing.