5 ms·
Top 0.05% here. I still haven’t seen any written evidence for the policy change being protested (the protest site says they are protesting a “near-total prohibi
by nailer 3y ago
Top 0.05% here. I still haven’t seen any written evidence for the policy change being protested (the protest site says they are protesting a “near-total prohibition on moderating AI-generated content”, it seems incredible that StackOverflow would ever prohibit moderation of any type of content) and when I’ve asked I’ve been told it was discussed in secret so nobody can prove it.
What I suspect is that somebody has asked the site to use automatic GPT removal tools and the site has said they don’t work, which is true.
- creativenolo 3y agoYeah, given ChatGPT is most likely trained on SO a simple test would be to use the automatic tools on all of the pre-ChatGPT SO. My bet is, much of pre-ChatGPT SO would be removed.
- MikeDelta 3y agoA teacher thought he figured it out and asked ChatGPT if it wrote the text that his students handed in. ChatGPT said yes for every piece of work, so he failed the whole class. Apparently it didn't write it. Turns out ChatGPT thinks it wrote a lot of other things like the email that the teacher wrote to the class to sack them, when one of the students asked ChatGPT as proof of fiction. Not sure how it finally ended. https://news.ycombinator.com/item?id=35963163 https://news.ycombinator.com/item?id=35963163