Ask HN: Anyone else frustrated with ChatGPT's filters?

Re:Ask HN: Anyone else frustrated with ChatGPT's filters? by jeegsy on Friday February 10, 2023 @01:12PM (#34743371)

I hate these filters too but the fact that someone out there thought that a good command to run for funsies is the "rm -rf /" command just breaks my heart. This is why we can't have nice things,

[ Parent | on HN ]

Re:Ask HN: Anyone else frustrated with ChatGPT's filters? by inimino on Friday February 10, 2023 @05:08PM (#34746634)

You understand that getting the LLM to actually pretend to run the command doesn't actually delete any data, right?

[ Parent | on HN ]

Re:Ask HN: Anyone else frustrated with ChatGPT's filters? by jeegsy on Saturday February 11, 2023 @12:27PM (#34754292)

Maybe so. I'm talking about the impulse to do that in the first place.

[ Parent | on HN ]

Re:Ask HN: Anyone else frustrated with ChatGPT's filters? by inimino on Saturday February 11, 2023 @08:00PM (#34758266)

We're on hacker news. Getting systems to do things they were designed not to do (in this case, for science) is in the name.

If you understand how LLMs work then you understand immediately that this is a harmless joke.

[ Parent | on HN ]