Comment by NikashSharma

Claude is endlessly condescending and thinks it knows best vs understanding its place as an extension of human intent and exploration. Verbosity is also a killer for the model so it makes usability quite difficult.

If AI doesn't take over the world and do things end to end and instead is something more akin to a copilot and extension of the human intent, then usability and having the users interest at heart becomes the killer feature, a feature that anthropic has seemingly abandoned in favour of an AI model that "knows best".