Skip to main content
Meta Stack Exchange

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

Required fields*

Cancel
45
  • 17
    Since this AI has been trained on data outside of the Stack Exchange network, will it ever provide non SO sources? It seems to bring up related SO posts when I ask questions, but obviously it is referencing other data as well. Makes me wonder what happens if it can answer a question without any SO references. Commented Jun 25 at 17:58
  • 82
    Looks like a thin wrapper on top of ChatGPT which doesn't necessarily have any connection to the site SO what-so-ever...? I can ask it completely off-topic questions or I can ask it programming questions. Just like I can do in any other GenAI prompt. Commented Jun 25 at 18:18
  • 162
    A huge portion of SE mods once went on strike over AI. Yet, SE Inc. shows again that it still hasn't understood to what extent generative AI is disliked by the community. You keep trying to force us to jump on the AI hype train that all the big companies have created even though it's been said countless times that the whole point of SO is that its answers have actual standards that generative AI doesn't (and can't) meet. If someone needs AI-made programming answers, they can go on ChatGPT, no need to duplicate it and slap a SO sticker on it. Commented Jun 25 at 20:12
  • 85
    What's the 'carbon footprint' for using this AI to find stuff on SO compared to a 'normal' search on SO? Is it better or worse for the environment/world? Commented Jun 26 at 8:03
  • 30
    This - and the related lies that "we never would" - seems to have aged like rotten milk. Thanks again for proving that my suspicions aren't about prejudice or antagonistic behavior... It is just reading the signs, even when white-knights keep telling you that "you are wrong". Commented Jun 26 at 8:08
  • 50
    No LLM on SO sites ! that's why I come here in the first place Commented Jun 26 at 8:58
  • 63
    @Rosie is the team also working on a fix for the nonexistent attribution, which in your own words is unethical? Commented Jun 27 at 12:46
  • 14
    @CPlus Frontend to ChatGPT. It says at the bottom "powered by OpenAI". Also the quality of the output is comparable to ChatGPT. Probably they generate the response than use it to search about it on SO and display questions. Not sure if they limit themselves to questions with accepted answers. I hope not. Commented Jun 28 at 13:52
  • 16
    @Rosie "Provided this early testing phase goes well..." I hope the responses to this post alone are clear in communicating the community doesn't want this on the site (regardless of how good you make it look on mobile) Commented Jun 29 at 15:10
  • 19
    In a world where I can ask any programming question and have at least 4 AI slop results that are semi-randomly distributed between "exactly what I needed" (though the first non-AI result would have been the same), "completely useless", and "total fiction"... this is a useless tool, unless leadership's only goal is to alienate their community and win funding from some investor who only knows how to read buzzwords (and, importantly, not what the actual users are saying) Commented Jul 3 at 17:55
  • 21
    @AshZade The fact that, after over a year, y'all are still trying to shoehorn AI into Stackoverflow, even after people opposed to it over and over again (the only reason to use Stackoverflow is the human to human Q&A), shows that our feedback is absolutely worthless. No amount of Product Manager speak will make any of us believe otherwise. Commented Jul 8 at 7:14
  • 22
    @AshZade ok, you've left out the most important part. As per your reply to b) the search results are not meant to be attribution, which means SE conciously released this without any attribution at all. It doesn't matter that you call this an alpha or an experiment. What matters is that SE promised not to do that, SE staff even wrote an answer to this very question which says attribution is required for ethical AI use, yet SE is still running an AI slop product without any concern for attribution. Who decided this for what reason, and why is it still live right now? Commented Jul 8 at 14:00
  • 21
    "We acknowledge the concerns around attribution of the chat responses, and we are working to address this in the best way for the community and users." Really, I wanted to go easy on the negativity. But seriously, what the? Attribution. Is. Nonnegotiable! This includes now, not when you finally figured out how to do this. Unless you are taking down this experiment until attribution is provided, you are not addressing this in the best way possible. Commented Jul 9 at 17:03
  • 9
    So you show on your own pages the output of a chatbot that was trained on our data and that never acknowledges anyone of us? Can you imagine how we feel? (Not acknowledged obviously but stronger terms come to mind.) Commented Jul 9 at 19:47
  • 10
    Do you ever look at all these heavily downvoted "brave new ideas" and wonder "maybe this is not what the community wants and we should rethink our direction"? Otherwise, why give people the possibility to express their opinion if it is not going to be respected anyway. Commented Jul 10 at 11:58