Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Large language models reduce public knowledge sharing on online Q&A platforms

academic.oup.com

Post reply on HN
Print view View original

81–90 of 366 posts

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#81

Post by wccrawford »

haolez wrote: post #72
PhilipRoman wrote: post #55

Earlier quoted context omitted.

I haven't really found stackoverflow to be that humiliating (compared to some IRC rooms or forums), basic questions get asked and answered all the time. But the worst part is when you want to do something off the beaten path. Q: how do I do thing X in C? A: Why do you need to know this? The C standard doesn't say anything about X. The answer will depend on your compiler and platform. Are you sure you want to do X ins…

The first time that I asked a question on #cpp @Freenode was a unique experience for my younger self. My message contained greetings and the question in the same message. I was banned immediately and the response from the mods was: - do not greet; we don't have time for that bullshit - do not use natural language questions; submit a test case and we will understand what you mean through your code - do not abbreviate…

A one week ban on the first message is clearly gatekeeping. What a bunch of jerks. A 1 hour ban would have been a lot more appropriate, and escalate from there if the person can't follow the rules.

Don't even get me started about how dumb rule 2 is, though. And rule 3 doesn't even work for normal English as many things are abbreviated, e.g. this example.

And of course, you didn't greet and wait, you just put a pleasantry in the same message. Jeez.

I'm 100% sure I'd never have gone back after that rude ban.

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#82

Post by atomic128 »

Eventually, large language models will be the end of open source. That's ok, just accept it.

Large language models are used to aggregate and interpolate intellectual property.

This is performed with no acknowledgement of authorship or lineage, with no attribution or citation.

In effect, the intellectual property used to train such models becomes anonymous common property.

The social rewards (e.g., credit, respect) that often motivate open source work are undermined.

That's how it ends.

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#83

Post by qntmfred »

rq1 wrote: post #61

People should just share their conversations with the LLMs online no? This would blogging 5.0. Or web 7.0.

That's pretty much what my youtube channel is turning into. just me talking to myself with chatgpt as co-host

eg https://www.youtube.com/watch?v=kB59Bz-F04E

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#84

Post by yapyap »

atomic128 wrote: post #82

Eventually, large language models will be the end of open source. That's ok, just accept it. Large language models are used to aggregate and interpolate intellectual property. This is performed with no acknowledgement of authorship or lineage, with no attribution or citation. In effect, the intellectual property used to train such models becomes anonymous common property. The social rewards (e.g., credit, respect) th…

no it won’t, it’ll just make it more niche than it already is.

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#85

Post by delichon »

tlogan wrote: post #67
melenaboija wrote: post #38

It’s been a relief to find a platform where I can ask questions without the fear of being humiliated Half joking, but I am pretty tired of SO pedantry.

The main issue with Stack Overflow (and similar public Q&A platforms) is that many contributors do not know what they do not know, leading to inaccurate answers. Additionally, these platforms tend to attract a fair amount of spam (self promotion etc) which can make it very hard to find high-quality responses.

I've gotten answers from OpenAI that were technically correct but quite horrible in the longer term. I've gotten the same kinds of answers on Stack Overflow, but there other people are eager to add the necessary feedback. I got the same feedback from an LLM but only because in that case I knew enough to ask for it.

Maybe we can get this multi-headed advantage back from LLMs by applying a team of divergent AIs to the same problem. I've had other occasions when OpenAI gave me crap that Claude corrected, and visa versa.

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#86

Post by appendix-rock »

mhh__ wrote: post #66
PhilipRoman wrote: post #55

Earlier quoted context omitted.

I haven't really found stackoverflow to be that humiliating (compared to some IRC rooms or forums), basic questions get asked and answered all the time. But the worst part is when you want to do something off the beaten path. Q: how do I do thing X in C? A: Why do you need to know this? The C standard doesn't say anything about X. The answer will depend on your compiler and platform. Are you sure you want to do X ins…

I find that this is mainly a problem in languages that attract "practical"/"best tool for the job" Philistines. Not going to name names right now but I had never really experienced this until I started using languages from a certain Washington based software company.

God. Yeah. I’ve always hated #IAMPolicies on Freenode :)

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#87

Post by verdverm »

tomrod wrote: post #73
verdverm wrote: post #69

Earlier quoted context omitted.

Plagiarizing means violating copyright, loosely speaking. When you, as a human, use SO, you assign your rights to the content to SO. That company is licensing the content to 3rd parties, including those who want to train their LLMs. What I find is that the LLMs are not spitting out SO text word for word, as one would when plagiarizing. Rather, the LLM uses the context and words of my question when answering, making t…

I thought plagarizing was producing new work substantially copied from prior work, regardless who owns the copyright? I thought this because self-plagarizing exists.

Well, if we could not reproduce with changes, what others have written and we have learned, it is unlikely we could make real progress. There are many more concepts, like fair use, meaningful changes, and other legalese; as well as how people use the term "plagiarize" differently. I never heard of this "self-plagarizing" concept, it seems like something fringe that would not be enforceable other than in the court of public opinion or the classroom via grades

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#88

Post by joshdavham »

verdverm wrote: post #64

For me, many of my questions about open source projects have moved to GitHub and Discord, so there is platform migration besides LLMs. I also tend to start with Gemini for more general programming things, because it will (1) answer in the terms of my problem instead of me having to visit multiple pages to piece it together, or (2) what it's wrong, I often get better jump off points when searching. Either way, LLMs sa…

> many of my questions about open source projects have moved to GitHub and Discord

Exact same experience here. Plus, being able to talk to maintainers directly has been great!

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#89

Post by GeoAtreides »

wccrawford wrote: post #81
haolez wrote: post #72

Earlier quoted context omitted.

The first time that I asked a question on #cpp @Freenode was a unique experience for my younger self. My message contained greetings and the question in the same message. I was banned immediately and the response from the mods was: - do not greet; we don't have time for that bullshit - do not use natural language questions; submit a test case and we will understand what you mean through your code - do not abbreviate…

A one week ban on the first message is clearly gatekeeping. What a bunch of jerks. A 1 hour ban would have been a lot more appropriate, and escalate from there if the person can't follow the rules. Don't even get me started about how dumb rule 2 is, though. And rule 3 doesn't even work for normal English as many things are abbreviated, e.g. this example. And of course, you didn't greet and wait, you just put a pleasa…

"I'm 100% sure I'd never have gone back after that rude ban."

upon saying this, the young apprentice was enlightened

Re: Large language models reduce public knowledge sharing on online Q&A platforms

#90

Post by appendix-rock »

LeadB wrote: post #57
melenaboija wrote: post #38

It’s been a relief to find a platform where I can ask questions without the fear of being humiliated Half joking, but I am pretty tired of SO pedantry.

For the major programming languages, it must be a pretty esoteric question if it does not have an answer yet. Increasingly, the free products of experts are stolen from them with the pretext that "users need to be protected". Entire open source projects are stolen by corporations and the experts are removed using the CoC wedge. Now SO answers are stolen because the experts are not trained like hotel receptionists (wh…

I’m sorry that you ran afoul of a CoC or whatever, but this sounds like a real ‘airing dirty laundry’ tangent.
Post reply on HN