A malicious response by users would be to employ an LLM instructed to write plausibly sounding but very wrong answers to historical and current questions, then an army of users upvoting the known wrong answer while downvoting accurate ones. This would poison the data I would think.
All use of generative AI (e.g., ChatGPT1 and other LLMs) is banned when posting content on Stack Overflow.
This includes “asking” the question to an AI generator then copy-pasting its output as well as using an AI generator to “reword” your answers.
Interestingly I see nothing in that policy that would dis-allow machine generated downvotes on proper answers and machine generated upvotes on incorrect ones. So even if LLMs are banned from posting questions or comments, looks like Stackoverflow is perfectly fine with bots voting.
A malicious response by users would be to employ an LLM instructed to write plausibly sounding but very wrong answers to historical and current questions, then an army of users upvoting the known wrong answer while downvoting accurate ones. This would poison the data I would think.
Ironic, isn’t it?
Interestingly I see nothing in that policy that would dis-allow machine generated downvotes on proper answers and machine generated upvotes on incorrect ones. So even if LLMs are banned from posting questions or comments, looks like Stackoverflow is perfectly fine with bots voting.
Sounds like it would require some significant resources to combat.
That said, that plan comes at a cost to presumably innocent users who will bark up the wrong trees.