kulkarniamey

31 ms·

1. 1.▲
   by kulkarniamey 18d ago

   Author here. I left in all those shipping changes to the agent - a prompt modification, a more lenient guardrail, some increased temperature -that altered the behavior of the agent without altering its execution, which flew past code review
2. 2.▲
   Show HN: Ctxwitch – Git tells you what changed; this tells you what it'll do(github.com)

   1 points by kulkarniamey 18d ago|1 comments
3. 3.▲
   by kulkarniamey 1mo ago

   The model is probably excellent. The problem here is AGI having various definitions and many of them getting narrowed down to whatever makes benchmark numbers look good.
4. 4.▲
   by kulkarniamey 1mo ago

   Just curious about agent mode - while comparing agent's code with that of human will you review is in same way ? for example: do you check if 'did AI actually understand the task ?' more that you would check for human changes