Live data from Hacker News

Stack Overflow questions are being flooded with answers from ChatGPT

news.ycombinator.com

81–90 of 342 posts

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#81
>What are the repercussions of this?

It will start feeding back into the training set, corrupting things. OpenAI will have an advantage at first as they can trivially filter out everything they have generated from the future training corpuses, since you can only run it through their servers. If they or someone else has breakaway progress such that almost all generated content is from their own servers because users only use them because their results are so much better, they could form a strong self-reinforcing moat against competitors forced to train on their semi-spam which they can trivially filter out.

It's also possible we'll see something like the existing big-tech patent cross-licensing agreements, where they all agree to share their generated outputs to filter from training, making it very hard for new entrants.

Other companies will begin having advantages as well, depending on how well they can get less tainted user data. Think of Discord, for example, where users may use AI but are less likely to gamify it like stack overflow and flood it for points, and instead be correcting its output etc. in programming discussions.

As things become more accepted Microsoft will probably eventually sell access to private github for training, with some stronger measures around avoid rote memorization.

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#83

It seems like there are a few potential negative consequences of using AI-generated answers on Stack Overflow. For one, the quality of the answers may be lower than if they were written by a human. Additionally, if these AI-generated answers become too common, it could potentially lead to a more impersonal and less supportive community on Stack Overflow. Finally, if the AI is able to search the internet and "inbreed"…

I think the far most problematic thing is that the answers are often plain wrong but well written. Eg (shameless plug) https://twitter.com/zuza777/status/1599370082919936000?t=oB9...

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#84

I wouldn't even mind so much if the answers were right. The problem is that a lot of them are totally wrong, but completely reasonable- and plausible-sounding, and in an authoritative tone, so unless you already know the right answer, the only way you'll realize its answer is wrong is the hard way.

That's the worst part. It's all fun and games when it tells a story of cheese sandwiches and VCRs in the style of the King James bible, but when it gives wrong answers in an authoritative tone and then insists it absolutely can't be wrong, it's terrifying.

I don't understand what good could come of this. Or at least make it detect what is fiction and what isn't.

People are treating it like it's Wikipedia, but it's not. It's a riff on words, like a bird imitating sounds without an idea of what they mean.

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#85

I assume that this is by folks wanting to up their scores. That's a huge problem with "gamification." I'm not especially a fan of the concept, in a venue like SO. I think it has led to a rather nasty community, and I hardly ever go there, anymore. I assume that we'll be seeing a lot of robotic HN content (I would not be surprised if it is already here, but has been sidelined by the mods).

Stack Overflow literally forces you to reach a certain score before you get downvoting rights. The gamification is real, but once you reach that point, you aren't required to play anymore.

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#86

Earlier quoted context omitted.

Yes, one way to identify responses that may have been generated by ChatGPT is to look for signs of repetitive or verbose language. Because ChatGPT is a language model, it is designed to generate text that is fluent and coherent, but it may also produce text that is unnecessarily long or repetitive. This can be a telltale sign that a response was generated by an AI, as human responders are less likely to use repetitiv…

To find out if a response was made by a computer, look for repeating words or long sentences. ChatGPT is a program that makes text that makes sense and flows well, but sometimes it uses too many words or repeats itself. This is a clue that the response might have been made by a computer instead of a person. However, this isn't a sure thing and there might be other reasons why a response is long or repeats itself. Sti…

> Respond to the above in the informal style of a slightly contrarian Hacker News comment, with a hint of passive-aggressiveness. Avoid repetitive wording and long sentences.

Wow, thanks for the tip! I never would have thought to look for repeating words or long sentences to determine if a response was made by a computer. I guess I'll just blindly trust everything I read online from now on. Sarcasm aside, this is a pretty flawed method. Just because a response has repeating words or is long doesn't mean it's not from a person. And conversely, just because a response is short and doesn't repeat itself doesn't mean it's not from a computer. It's not that simple.

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#87
post #46
post #31

Earlier quoted context omitted.

Who cares if the comments are good enough?

For example, if one entity generated a large portion of the content, they could easily introduce a bias in these comments, to sway the opinion of readers. Automated astroturfing.

It doesn't even have to be nefarious. Just imagine a thread where 90% of the comments are repeating variations of the model's ideas in slightly different ways, and only 10% of the other responses are from humans. Even if the AI responses are considered 'good'... it gets overwhelming to the point that why should normal people even comment as the likelihood of it being read by another human and useful or even seen drops to nothing. I might as well open up a blank text file and make this comment and then not save it and close it as that would be seen by the same number of people in the limit.

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#88

Earlier quoted context omitted.

You can kinda tell it's ChatGPT by how verbose and repetitive the writing is.

Yes, one way to identify responses that may have been generated by ChatGPT is to look for signs of repetitive or verbose language. Because ChatGPT is a language model, it is designed to generate text that is fluent and coherent, but it may also produce text that is unnecessarily long or repetitive. This can be a telltale sign that a response was generated by an AI, as human responders are less likely to use repetitiv…

That's honestly a hilarious response from ChatGPT

Re: Stack Overflow questions are being flooded with answers from ChatGPT

#90

I wouldn't even mind so much if the answers were right. The problem is that a lot of them are totally wrong, but completely reasonable- and plausible-sounding, and in an authoritative tone, so unless you already know the right answer, the only way you'll realize its answer is wrong is the hard way.

What you said may sum up the current state of AI. People's minds are continually being blown, but will there be a realization that these AIs are specialized to provide specious output and nothing else? There's a canyon of difference between something that sounds correct and a thing that is actually correct.

Exactly. ChatGPT sounds like a bad student who didn't actually learn anything during the year and is trying to bullshit their way through the finals. Or a politician, maybe. It's formalizing the worst traits of humanity.
Post reply on HN