Q & A site Stack Overflow has changed its policy to 'AI posting OK' and is controversial, more than 600 moderators signed a large-scale strike to protest
Volunteer moderators active in Stack Overflow, a knowledge community about information technology and programming technology, announced an open letter to Stack Exchange, which operates Stack Overflow, on June 5, 2023, and volunteered for community management and maintenance. announced a pause. In a letter, the moderators called for the withdrawal of the new policy, which in principle prohibits the immediate deletion of AI-generated posts.
Dear Stack Overflow, Inc.
Until now, Stack Overflow has had a policy of banning AI-generated posts such as ChatGPT, and when moderators discover such posts, they will immediately delete them, or in some cases bypass the normal process. It was possible to immediately freeze the poster's account by doing so.
Behind this measure, AI-generated posts contain seemingly correct but actually incorrect content, so-called ' hallucinations ', which are difficult for general users who are not skilled in coding to properly distinguish. Therefore, there is concern that traditional features such as voting systems will not be able to keep up with the deluge of inaccurate posts.
Conversation AI 'ChatGPT' answer posting is temporarily banned on Stack Overflow, a coding Q & A site-GIGAZINE
However, on May 29, which was a public holiday for many moderators living in the United States and the United Kingdom, an opinion was posted at a private moderator meeting that questioned the policy of uniformly deleting posts of AI-generated content.
In a post on Meta Stack Exchange, a site that discusses Stack Overflow's management policy, etc., with the contents partially changed on the following day, 30th, ``Detection of AI-generated content is inaccurate and may cause false positives. It contains a claim that the number of posts is very high,” in which the poster asks to stop deleting posts solely because they were generated by AI.
Additionally, the post was tagged 'mod-agreement-policy'. This resulted in a binding moderator policy, treating the policy of not removing AI-generated content as agreed upon.
In response to this incident, the moderator who carried out the strike posted on Meta Stack Exchange, 'The new policy ignores the established community consensus and the support of content moderators so far, and community members It is not discussed at all, is presented in a way that is misleading to moderators and users, and is based on unsubstantiated claims derived from data analysis that has not been reviewed and cannot be reviewed.' I am convinced that allowing AI-generated content disguised as generated content will eventually reduce the value of the site to zero.'
AI-generated content isn't the only problem. Moderators claim that Stack Exchange has repeatedly forced changes that moderators objected to, and that even requests for improvements were sloppy. For example, chat, which is the most necessary tool for moderators and curators, is a very old feature, so the community has repeatedly requested improvements, but it has been neglected for years.
Due to these concerns and distrust of the operating company, the moderators wrote an open letter, saying, ``Our efforts and concerns to bring about change through appropriate channels have been ignored at every turn. As a last resort, we will discontinue our dedication to the platform that has devoted volunteer efforts for over 10 years, ”he declared the strike.
The specific activities covered by the strike are as follows:
• Flag and handle problematic posts.
・Run anti-spam bot 'SmokeDetector'.
- Closing a post or voting for its closure.
- Delete posts or vote for deletion.
- Review various tasks that are added to the review queue.
- Running various bots designed to aid moderation, such as detecting plagiarism, low-quality answers, and rude comments.
Regular users who are not moderators or curators can also participate in a strike by refraining from voting or commenting on posts. However, since we are aiming for a moderate strike, we ask that you refrain from intentionally posting AI-generated content or performing vandalism.
The duration of the strike will be indefinite, and will allow Stack Exchange to 'withdraw the policy change in question in a way that addresses some of the concerns of moderators and allows moderators to effectively enforce responses to AI-generated answers.' until ”.
The letter had more than 600 signatures at the time of writing, including moderators and general users.
Related Posts:
in Software, Web Service, Posted by log1l_ks