Biggest utility I see for AI personal assistants is in a recommendation system based on personal preferences. We are not breaking any barriers with this. Trust

Biggest utility I see for AI personal assistants is in a recommendation system based on personal preferences. We are not breaking any barriers with this. Trust implies some sense of consistency in results and LLMs have proven to be the opposite. Your whole argument's premise seems to be built on some of the dazzle created by AI assistants directly translating to advertising value. The connection is very loose here and to me only propagates the hype that's all too common in discussions about AI.