A virtual war between the administration and users of the popular online forum for programmers Stack Overflow has broken out. The reason for this was Stack Overflow’s partnership with OpenAI to train artificial intelligence models based on the forum users’ posts.
Not everyone liked it. After the announcement of this partnership, many Stack Overflow users started deleting or modifying their questions and answers. By doing so, they wanted to prevent their materials from being used to train ChatGPT.
However, this behavior of users did not please the administration of Stack Overflow. As a result, the forum moderators began to punish users who resort to deleting or editing their previous posts by blocking them. While users say that they have the right to remove their own content from the forum based on the «right to be forgotten», the Stack Overflow administration refers to the forum’s terms of use (which everyone agrees to follow, but almost never reads when registering). The terms contain a clause about the irrevocable right to own all content.
Users who disagree with their content being used for ChatGPT training are particularly outraged by the rapid changes in Stack Overflow’s policy on generative AI. For years, the site had a standing rule that prohibited the use of generative AI to write or reformulate any questions or answers. Moderators were allowed and encouraged to use AI detection software when reviewing posts. However, the company has recently changed its AI policy.
Stack Overflow user Ben posted a story on Mastodon about his experience of editing his most successful answers to try to avoid his work being used for OpenAI purposes. Among other things, he wrote:
«[Moderator repression] is just a reminder that anything you post on any of these platforms can and will be used for profit. It’s only a matter of time before all of your Discord, Twitter, etc. posts are collected, fed into a model, and sold back to you».
Source: tomshardware