OpenAI says that its latest technology, GPT-4, which powers its ChatGPT software, can be used to radically streamline and speed up content moderation practices.
The AI developer — which says that its tools can be used to develop content policies and label and make decisions about posts exponentially more quickly than current, time-intensive moderation techniques — is testing the application and encouraging customers to experiment with it, according to a Bloomberg News report.
“A content moderation system using GPT-4 results in much faster iteration on policy changes, reducing the cycle from months to hours,” OpenAI states in its corporate blog. “GPT-4 is also able to interpret rules and nuances in long content policy documentation and adapt instantly to policy updates, resulting in more consistent labeling. We believe this offers a more positive vision of the future of digital platforms, where AI can help moderate online traffic according to platform-specific policy and relieve the mental burden of a large number of human moderators. Anyone with OpenAI API access can implement this approach to create their own AI-assisted moderation system."
Meta and other tech giants have struggled with trying to monitor and curb the most dangerous excesses on social media. That challenge has been complicated by political polarization, with some politicians urging more control to reduce widespread misinformation undermining democratic institutions and others trying to stop all moderation, claiming First Amendment grounds.
By compressing what is now months’ worth of work down to hours’ worth, ChatGPT-4-based applications could free platform employees and others up to “focus on more complicated decisions related to the most extreme cases of potential content violations and how to refine policies,” Andrea Vallone, a policy official at OpenAI, told Bloomberg.
In addition, it could spare human content moderation employees from having to review huge numbers of hate posts and disturbing images.