Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

For what it's worth, I've not yet met an LLM that both produces coherent text and will talk past you & try to get under your skin. I'm sure it could be done with an open source model and elbow grease but my read on the situation is that there's a person on the other side of the screen, who for whatever reason had some kind of mental block around this subject matter.

Oh well. Wishing you unachievably bold aspirations. To paraphrase the Tao Te Ching, the path that is clear does not go all the way.



Yeah that's kinda what got me too. But whatever it writing those responses (troll or machine) is becoming less coherent as we talk more. I'm not sure if LLM but it does remind me of that part where they get stuck on something and will correct for a split second before going back to whatever they were doing. But I'm not sure, it's just weird and I'm certain the writer is not being genuine

I'm not sure why they're trying to manufacture outrage but I've been seeing interactions like that appear more and more on HN.


"They're" right here in the comments section with you and simply a person who disagrees with you.

Calling someone names and making baseless accusations is a good sign that you're out of anything meaningful to add.

> I've been seeing interactions like that appear more and more on HN.

Is continuing to compare other commenters to LLMs an example of the type of helpful and beneficial interaction you'd like to see in comparison?


  > you're out of anything meaningful to add.
You're right. I think that was mentioned a number of comments ago. We haven't been shy about noting that the conversation is going in circles and you've failed to address our points or present alternatives. You have absolute confidence in your solution, yet have not elaborated on what that is or how it solves the problems we've discussed. You've continued to hyperfixate on an alias. You've continued to complain about an inability to do things a certain way while we've stated that your way also works. I mean that's what an alias is...

The problem isn't that we disagree with you, the problem is you are not making an attempt to have a productive conversation. So we gave up.


> "we"

Was there a meeting where you were made spokesperson for you and this other person, or did you just start speaking for them?

> So we gave up.

What part of responding nonstop in great detail with intricate formatting indicates you've given up, even if just speaking for yourself this time?


This isn't about bots or mental blocks. It's about you needing everyone who disagrees with you to be defective somehow—artificial, mentally blocked, aesthetically driven, anything but right.

Your fortune cookie sign-off does no justice to the work you claim to be referencing. I guess if you spend your days working with agreeable LLMs, it is very inconvenient and upsetting when a person comes along who doesn't agree and polish up every concept you espouse.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: