If a discussion about GenAI adds something new to the discussion, I think it is great. But most of them do not. Most of them are either generic posts about whether or not someone likes it, stories of someone's experiences as they learn it, or fanfiction about where they want it to go in the future. Or, at the worst, links to simplistic wrapper apps someone is trying to market.
I would prefer that it lands in the same place as any other topic - if it is interesting, it gets traction, and if it is boring, it does not. But how to make that happen is a challenge.
I was thinking of making an alternative frontend to HN with GenAI stuff filtered out — using GenAI to do it, obviously (both build it and use it to filter things).
But mostly as a social experiment to test how fed up are people with AI-in-our-face ;)
This is what my sentiment is: there is genuinely useful stuff, but it really overwhelmes anything else and it's hard to find non-AI valuable stuff on the frontpage these days.
I do apprehend this kind of ideas. It would make it easier for models to be trained to bypass that filter, making it even harder for humans to detect slop.
IMO this would be like in 1997 (or pick whatever year you think the Internet took off) asking, "How do you feel about posts about the Internet taking over the HN front page?"
It is kind of annoying but to be fair, developing software with agents is a big paradigm shift in the field so its not surprising.
HN is mostly dead to me. The site was billed as a place for curious minds, but there is little curiosity in letting a machine do your thinking.
If a discussion about GenAI adds something new to the discussion, I think it is great. But most of them do not. Most of them are either generic posts about whether or not someone likes it, stories of someone's experiences as they learn it, or fanfiction about where they want it to go in the future. Or, at the worst, links to simplistic wrapper apps someone is trying to market.
I would prefer that it lands in the same place as any other topic - if it is interesting, it gets traction, and if it is boring, it does not. But how to make that happen is a challenge.
I was thinking of making an alternative frontend to HN with GenAI stuff filtered out — using GenAI to do it, obviously (both build it and use it to filter things).
But mostly as a social experiment to test how fed up are people with AI-in-our-face ;)
This is what my sentiment is: there is genuinely useful stuff, but it really overwhelmes anything else and it's hard to find non-AI valuable stuff on the frontpage these days.
Thank you. It is good to know that I am not the only one feeling this way.
I do apprehend this kind of ideas. It would make it easier for models to be trained to bypass that filter, making it even harder for humans to detect slop.
This is not about detecting AI generated content but detecting HN posts which mention/discuss AI.
Fun fact but I suggested it by making a Ask HN post about it nearly 2 years ago at this point:
Ask HN: Are you interested in a Hacker News alternative which doesnt focus on AI:https://news.ycombinator.com/item?id=41789661 (oct 9,2024)
I for one would be interested if you do end up making one.
Some people have ended up making Ublock origin blocklists which block any post which mentions AI iirc.
I am not surprised others had a similar idea well before me: I'd just keep it as a "lens" into HN, and not a new thing.
As always, it's all about execution: we just need someone to do it.
> As always, it's all about execution: we just need someone to do it.
Agreed.
Well, it sure is a relief from all of the Rust Über alles posts. There will always be trends, this is just the latest.
Money got expensive which is why everyone is being let go. AI is a convenient cover story to deflect attention.
I feel awesome about it!
IMO this would be like in 1997 (or pick whatever year you think the Internet took off) asking, "How do you feel about posts about the Internet taking over the HN front page?"