zlacker

The comment section on HN is for comments and discussions of people so yes any form of generated responses does IMHO fundamentally disqualify.

I mean to nitpick if a AI becomes sentient and flexible enough to be called a person and decides to comment on their own that would be fine.

But when it comes to:

- auto generating responses

- hiring people to generate responses for you

- anything similar to the other points

I think it should not be allowed.

But I have no idea how to enforce this reasonably.

>>dathin+(OP)
Yeah the last point is quite crucial. Should we ban things that are hard to enforce? Does that lead to selective enforcement by looking into responses you don't like?

>>dathin+(OP)
I have a disability. I'm a bad writer and a bad speller. It takes me a long time to compose a response that I wont later regret.

I haven't looked into it yet, but I was thinking ChatGPT might help me. I could quickly jot down my thoughts, and let the computer arrange them in an easier to read format.

>>jay_ky+We
What I mean is not allowing responses of which the content is auto generated not that the text syntax/prosa is auto formulated.

I don't believe any form of auto detection of ChatGPT-like text will long term work, even through it has a certain style you can change that style or post process it and that's just the first version of ChatGPT.

>>jay_ky+We
That's not what ChatGPT does though. It will always add its own "thoughts" to your prompts. Using it as a writing assistant for spelling and grammar won't work.

>>onion2+SA1
That's not true. It has refined the texts I gave it without adding "its own thoughts" many times for me.

Even so, you can easily edit anything undesired out.