Yeah i am so sick and tired. It was so refreshing to read HN and always find something interesting, now its tiresome, AI marketing everywhere and comments from people that doesn't even like to program and they now can develop a todo list app at the expense of 200$ for whatever claude has release that week.. sad.
I've always thought it would be a sorta neat project to make a web plugin to remove certain topics from the HN main page, but I have never before been ->this<- close to starting such a project than I am now. These topics have totally taken over, and quite frankly, the volume of them make HN less interesting.
I wrote a simple uBlock Origin rule as a quick-and-dirty solution months ago. It doesn't get rid of sibling nodes (still haven't researched how to do it), but does remove the submission title and the vote button.
Bonus points if you implement it with an LLM, then blog about it using an LLM, and then post the link to HN where it will be auto filtered out by the generated tool (maybe, because let’s face it, it’s non deterministic)
Anything that gets posted to HN when you’re born is normal and ordinary and is just a natural part of the way the world works.
Anything that gets posted to HN between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
Anything posted to HN after you're thirty-five is against the natural order of things.
The GP wasn’t complaining about there being an AI hype. They were complaining that the front page is oversaturated by said hype.
I’ve been on HN since near the beginning (under a different user name originally). And there have been quite a few trends rise and fall on here. But none of them were as intense as this AI hype currently is.
In fact it’s not even AI in the more general sense, it’s almost entirely just LLMs that get discussed. There’s so much more going on outside of LLMs but all anyone is talking about on HN is natural language tools.
> In fact it’s not even AI in the more general sense, it’s almost entirely just LLMs that get discussed.
"AGI is right around the corner" "No it's not" "Yes it is, LLMs are the future." "We don't even know if AGI is possible." "LLMs are the future." "No they aren't." "AGI is right around the corner..."
or
"LLMs are really useful." "No they're not" "Yes they are." "No they aren't." with a little bit of "They sucked the last time I used them." "Did you use them recently?" "That's what someone said last time." "But LLMs are really useful" ...
over and over and over.
It isn't even that it's mostly just LLMs being discussed, it's how they're begin discussed, they're effectively just a proxy for optimists and pessimists to argue over which worldview is better.
If we were talking about AI in general and not LLMs, the same conversation structures would still pop up.
No because those conversations are specific to off the shelf models used in generalised ways, which basically only applies to LLMs.
When you look at other applications of AI, such as models trained for medical usage or AI used in Hollywood (something I personally have professional experience in) then it’s a completely different story because they’re bespoke models trained and used for specialised edge cases. Rather than Joe Blogs using an off the shelf package to do the same thing as the previous person albeit slightly differently.
I imagine that's because LLMs are of most interest to the Hacker News crowd: they can help write code, and you can build systems on top of them that can "understand" and respond in human language.
Generative image / video / audio models can produce output in image, video and audio. Those have far less applications than models that can output text, structured data and code.
HN is by ICs who write code, the 90% of the folks that build all the stuff, largely neutral to negative. It has gained some excellent traction with 10% of the folks, but it is quite behind compared to ai coding subreddits. Months behind.
In a kind of personal plot twist way I'm happy with the outcome. I've now spent way less time here.
If before there was a bit of an attachment to checking what was going on on HN, now there's an overall meh. I still do visit out of habit, but when I do, I have this blasé attitude that quickly takes me away. Too much LLM this or that, AI everywhere.
HN being one of the last Social Media sites I've engaged with, it's good to finally let it go. Yay?
>Dunno, I felt way worse during the crypto boom days. At least the AI stuff has a bit more of a generally useful application.
The problem with the current wave of AI is that it isn't "generally useful" yet. These AI systems can be very effective in certain situations when they are used with the specific knowledge of their flaws, but they are being applied to a much too broad use case today. Granted the valid use cases for crypto were miniscule in comparison, so I won't quibble too much with your general point.
Although the most frustrating aspect to me is that the people who were shown to be right about crypto have not earned any gravitas when they say similar things about AI. So many charlatans were saved by a second bubble immediately inflating as the previous bubble popped and we are once again ignoring the people calling them charlatans despite the proven track record for calling out this behavior. Seemingly no one learned anything.
I mean, some of us are still uploading various hacker-newsy things like this article by Naughty Dog on how they used a custom LISP for their development of Crash Bandacoot for the Sony PlayStation — but they get like, one upvote ;)
That's an old and fairly well-known thing, though. Lisp folks don't see a lot of commercial success stories so we tend to talk about the ones we have. The non-Lispers don't care.
