Your last chat with an AI holds the wreckage of a scorecard. Here’s how to recover it.

1 points•grrrtttt•10 days ago•1 comment•

1 comment

verdverm10 days ago
the space of negative clarifications is infinite

the problem is the training, to be relentless spaghetti throwers until the "think" the task is complete

train / fine-tune them to stop more and ask questions

Read the full thread on Hacker News →

Related stories