I have a lot of doubts that this would be very useful. LLMs still have a lot to prove in many respects of reliability, insight, energy usage, context awareness. This feature would be a mere gimmick against other features more conventionally developed...
