4 ms·
Once you've seen the AI not make mistakes like that a few dozen times you start to trust it not to mess that up in the future.
by simonw 3mo ago
Once you've seen the AI not make mistakes like that a few dozen times you start to trust it not to mess that up in the future.
- discreteevent 3mo agoIs this really a rational strategy for something whose nature is to be right most of the time and then spectacularly wrong a much lesser amount of the time?
- simonw 3mo agoYes, because the mistakes in code are easy to spot. I wouldn't use this to write me an unreviewed legal brief.
- discreteevent 3mo agoBut you said that in this case you don't review the code because you trust the AI having seen it write that kind of code before.
- twister2920 3mo agoyou literally just said you don't review this code!