Moderators on the popular programming forum Stack Overflow have gone on strike in protest of the site's decision to prevent the moderation of AI-generated content. The moderators argue that the new policy will allow low-quality and misleading content to proliferate on the site and that it will make it more difficult for users to find accurate information.
In an open letter, they accuse the operator company of the code-help platform and related forums of Stack Exchange of “imposing a near-complete ban on moderating AI-generated content”. AI-generated content is actually prohibited on the platforms.
The strike was triggered by a new moderation policy that states that the moderators can no longer use the assumptions they have made so far to make decisions. This includes the writing style of an answer or the general behavior of a contributor, from which the moderators previously concluded whether the content was created by an AI.
The moderators and users are particularly concerned about ChatGPT and similar AI bots that generate code snippets or answers based on natural language queries. These bots are often used by inexperienced programmers who want to get quick solutions without understanding the underlying concepts or logic.
ChatGPT and Similar Chatbots are Banned
The problem with these bots is that they can produce incorrect, insecure, or even malicious code that can harm other users or systems. Moreover, they can also violate the intellectual property rights of the original authors of the code snippets or answers.
The moderators and users argue that ChatGPT and similar AI bots are banned according to the terms of service of Stackoverflow and Stackexchange, which state that “you must not use any robot, spider, scraper, or other automated means to access this site for any purpose”. They also point out that these bots can damage the reputation and trustworthiness of the platforms as sources of reliable and high-quality information.
You may remember that Stack Overflow took the decision to ban ChatGPT last December. According to mods on Stack Overflow, the ban was put in place because ChatGPT makes it too easy for users to generate AI responses and essentially spam the site with responses. Many of these answers look helpful but end up being incorrect.
Mods say the ban is currently temporary but a final ruling will eventually be made about the ongoing support of OpenAI's chatbot:
“The primary problem is that while the answers which ChatGPT produces have a high rate of being incorrect, they typically look like they might be good and the answers are very easy to produce,” claim the mods.
ChatGPT is Overtaking Stack Overflow
We recently reported that ChatGPT has been attracting more users at the expense of Stack Overflow, the leading online forum for programming questions. Experts say that ChatGPT's ability to provide programming assistance, fix code errors, and respond to complex queries makes it a convenient alternative for programmers who want fast solutions. David Carr, a senior insights manager at Similar Web, wrote in a blog post that “ChatGPT reduces the need for programmers to ask for help on platforms like Stack Overflow”.
Stack Overflow is a platform where coders and programmers can ask and answer questions about various programming topics. They can also use the source code that is attached to the questions and answers on the site to copy or learn from (parts of) it.
Instead of seeking advice from other developers on Stack Overflow, many programmers seem to prefer to turn to ChatGPT first. The underlying GPT-3.5/GPT-4 language model from OpenAI was trained on massive web data which supposedly also includes online spaces such Stack Overflow, blogs and coding communities.