← Back to context

Comment by mlrtime

21 hours ago

I think it's human nature to point at something you don't like and if it isn't 100% perfect then point to it and say it's flawed and must be taken down.

Repeated examples on HN

- TSA effectiveness

- AI Writing code free bug

- Self driving cars get into accidents

You are missing an important element. You can decide for yourself whether AI-produced code is worth the price. You don't get to decide whether the TSA is effective enough to pay for it.

Maybe you are willing to pay 15% for AI that saves you 20%. Even if it isn't very effective, you come out slightly ahead. Or maybe you pay 85% for something you deem to be 90% effective

With TSA you pay 300% for something you might judge to be 2% effective and you don't have a choice.

- TSA fails its own Red Team exercises 95% of the time.

- Self driving cars have measurably fewer accidents.

If you're confusing the two, I suggest you look into the data.

*Not sure on AI code yet.