The opposite may be true, the more effective the model the lazier the prompting as it can seemingly handle not being micromanaged as with earlier versions.