Idle Sunday thoughts on why so many people love LLMs for 'editing' even though it gives that 'je ne sais qua' of corporate blandness. I'd assumed it was the perceptual con: chatbots keep making generic statements that users interpret as specific
It works a bit like a "soup stone". You're the one providing the work, observation, and analysis—the ingredients—while the credit goes to the 'soup stone'. Except the stone is charging you Michelin star prices for the soup
