[diatone]

Yeah it’s bizarre; if someone’s going to get an LLM to ghostwrite their reply (not saying this is the case here but my internal ChatGPT alarm also went off reading this), will they use it for replying in the thread? Will they fact check their work?

[squeaky-clean]

I didn't think to check their profile before since it's not a green account, but they have no comments in the past 67 days and now 2 posts in 12 hours promoting their AI video translation service before being active again. The writing style in all their other comments feels different. So I'm pretty confident now this is just chatgpt.

[philomath_mn]

The complete, unquestioning acceptance of the term "ScaleUp CTO" also stuck me as something GPT would do

[souvic]

Well, I provided that term to ChatGPT, not the whole post and also I have read at least 6 of these 10 recommendations, modified the list as per my knowledge and then posted it here. I think ChatGPT can and definitely be used for re-drafting purposes. What's wrong with that?

It is still my original response and I have read most of the books with few being in my wishlist.

[squeaky-clean]

Well firstly that means you're recommending 4 books you haven't personally read without saying they came from chatgpt (and we don't know which 4).

But also the writing style in that comment just feels far more like ChatGPT than someone elses. Like you asked ChatGPT and then edited that, and not that you came up with your own idea and asked chatgpt for help. So when reading the comment it just comes off as any other copy-paste ChatGPT response. The OP could just ask ChatGPT themselves.

It also doesn't scale well. If multiple people in this thread used ChatGPT to answer, we'd have a bunch of comments of mostly the same list. When I tried to asked it myself I got 7 out of 10 recommendations the same.

[philomath_mn]

I come here to hear what humans have to say. I'm all for using GPT to make life more efficient, but using it to write your HN posts misses the whole point of HN.