Discussion is much more focused here and not that much actually occurs. E.g. try following https://news.ycombinator.com/newcomments and most of the time it's not difficult to read through every single comment being made. If what you're looking for is well discussed on HN regularly though then it's great.
There's a lot of good troubleshooting and tech info on there, though. I've run into multiple tech issues over the years that aren't documented in bug reports, change logs, etc., but some random person on Reddit will have a post where they investigated the cause and found a solution.
For software, if it already has the cloned source then it can just look at how the thing works directly and answer you completely. I should think they'd already be cloning all FOSS repos and that it'll eventually just do that.
And quite often a thread isn't just a problem and a solution but the set of steps to get to the destination. Even if it's not the exact same issue the process is useful.
It’s also just unreliable; people complain about Wikipedia being user edited but Reddit is so often just nonsense. It’s what the collective wants to be real, not what’s actually real
I don't know why this matters at all. "Citations" are just RAG ex post facto bullshit excuses. They're not real citations for what the LLM ingested, and they are not associated with any particular chain of tokens they can output.
If LLMs have ingested Reddit then they've ingested Reddit. That doesn't need to change. They will continue to be confidently stupid and haplessly helpful, in reverent homage to every Reddit incel inhabiting MOm's basement with a lifetime supply of Snickers and Dr. Pepper.
They can certainly still ingest and rehash Reddit, without RAGing them and citing links.
This really is the dumbest timeline. A few decades ago, futurists envisioned semi-autonomous agents on the 'net, like obedient golden retrievers, who would seek out your favorite headlines and articles, and collate them for you to read like a customized morning paper. Instead we now have clickbait farms, idiot forums controlled by the GNAA, and the lamestream media continuing their ridiculous nightly distractions from reality.
Reddit started charging for access: https://arstechnica.com/ai/2024/02/reddit-has-already-booked...
That’s from 2024
I'm aware. The age of the data is likely why they stopped showing results.
Good riddance
Hard to see this as anything but a good thing, Reddit is such a cesspool.
It's still one of the rare places where public discussion is happening. There is simply no alternative.
Is this entire site not an alternative to a sub-reddit?
Discussion is much more focused here and not that much actually occurs. E.g. try following https://news.ycombinator.com/newcomments and most of the time it's not difficult to read through every single comment being made. If what you're looking for is well discussed on HN regularly though then it's great.
...like a sub-reddit
The public discussion on Reddit is very heavily censored and probably already dominated by bots at this point.
Reddit is like 98% bots and 2% mods at this point
There's a lot of good troubleshooting and tech info on there, though. I've run into multiple tech issues over the years that aren't documented in bug reports, change logs, etc., but some random person on Reddit will have a post where they investigated the cause and found a solution.
For software, if it already has the cloned source then it can just look at how the thing works directly and answer you completely. I should think they'd already be cloning all FOSS repos and that it'll eventually just do that.
And quite often a thread isn't just a problem and a solution but the set of steps to get to the destination. Even if it's not the exact same issue the process is useful.
This is most likely a good thing for both platforms
It’s also just unreliable; people complain about Wikipedia being user edited but Reddit is so often just nonsense. It’s what the collective wants to be real, not what’s actually real
A very small, specific and highly correlated slice of the collective, though.
[dead]
I don't know why this matters at all. "Citations" are just RAG ex post facto bullshit excuses. They're not real citations for what the LLM ingested, and they are not associated with any particular chain of tokens they can output.
If LLMs have ingested Reddit then they've ingested Reddit. That doesn't need to change. They will continue to be confidently stupid and haplessly helpful, in reverent homage to every Reddit incel inhabiting MOm's basement with a lifetime supply of Snickers and Dr. Pepper.
They can certainly still ingest and rehash Reddit, without RAGing them and citing links.
This really is the dumbest timeline. A few decades ago, futurists envisioned semi-autonomous agents on the 'net, like obedient golden retrievers, who would seek out your favorite headlines and articles, and collate them for you to read like a customized morning paper. Instead we now have clickbait farms, idiot forums controlled by the GNAA, and the lamestream media continuing their ridiculous nightly distractions from reality.