3 ms·

Enough with the ChatGPT already. If you didn't write it, don't post it.

Grimburger 3y ago
There really should be some law or adage coined for this by now. No one wants to hear GPT's opinion because it doesn't have one.
- TeMPOraL 3y ago
  Nah, GPT's output can be very valuable and insightful. But you don't get it with a low effort "please summarize the key points of the text below\n\n{paste some text here}". I propose the rule to be what it is already: put some effort into your comment, make it add something instead of just taking space and attention. This works for human and GPT and human/GPT ensemble commentary. And the prescription for low-effort spam? Downvote and flag.
- wiseowise 3y ago
  I do. Sometimes.
- johnnyanmac 3y ago
  >No one wants to hear GPT's opinion because it doesn't have one. google doesn't have an opinion either when taking any search result at large. And that's all ChatGPT is doing here.
esquivalience 3y ago
I thought the approach was interesting. It's being used here as an empathy engine: what might (other people) say?
- bredren 3y ago
  Thanks. Most everyone seems to have missed the point, of the copy paste. My goal was to show how trivial it would have been for the CEO to have typed that into chat gpt and avoided the injury of unforced errors. Sort of a modern “let me google that for you.” I had this written out for the comment but at the time couldn’t make my explanation add more to the simple provision of “what you get when you try this using present tech” and a simple prompt and response. Another way to look at it is that LLM queries and their responses serve almost like ultra niche links to the docs except their responses vary and may even be transient. Providing a simple one could be read like RTFM.
buildbot 3y ago
I’d support instant bans for drivel like this, it’s somehow way more annoying than low effort/meme comments. Like, we get it, you can type shit into a chat interface hooked up to a LLM. So can we all.
- account42 3y ago
  Same for all the "what if we add an LLM to this" comments. Worse than the unispired blockchain "problem solving".
bowsamic 3y ago
I can't believe how much ChatGPT responses are being posted to HN unashamedly. It's really bizarre and useless

There really should be some law or adage coined for this by now. No one wants to hear GPT's opinion because it doesn't have one.

TeMPOraL 3y ago
Nah, GPT's output can be very valuable and insightful. But you don't get it with a low effort "please summarize the key points of the text below\n\n{paste some text here}". I propose the rule to be what it is already: put some effort into your comment, make it add something instead of just taking space and attention. This works for human and GPT and human/GPT ensemble commentary. And the prescription for low-effort spam? Downvote and flag.
wiseowise 3y ago
I do. Sometimes.
johnnyanmac 3y ago
>No one wants to hear GPT's opinion because it doesn't have one. google doesn't have an opinion either when taking any search result at large. And that's all ChatGPT is doing here.

esquivalience 3y ago

I thought the approach was interesting. It's being used here as an empathy engine: what might (other people) say?

bredren 3y ago
Thanks. Most everyone seems to have missed the point, of the copy paste. My goal was to show how trivial it would have been for the CEO to have typed that into chat gpt and avoided the injury of unforced errors. Sort of a modern “let me google that for you.” I had this written out for the comment but at the time couldn’t make my explanation add more to the simple provision of “what you get when you try this using present tech” and a simple prompt and response. Another way to look at it is that LLM queries and their responses serve almost like ultra niche links to the docs except their responses vary and may even be transient. Providing a simple one could be read like RTFM.

buildbot 3y ago

I’d support instant bans for drivel like this, it’s somehow way more annoying than low effort/meme comments. Like, we get it, you can type shit into a chat interface hooked up to a LLM. So can we all.

account42 3y ago
Same for all the "what if we add an LLM to this" comments. Worse than the unispired blockchain "problem solving".

bowsamic 3y ago

I can't believe how much ChatGPT responses are being posted to HN unashamedly. It's really bizarre and useless