ChatGPT: A Potential Moderator For Content on Social Media Platforms

The potential integration of GPT-4, a prominent Large Language Model utilized by ChatGPT Plus, into the function of an online moderator is being considered. This would involve the task of monitoring forums and social networks to identify and address objectionable content that is deemed inappropriate for public consumption. According to a recent blog by OpenAI, the developer of ChatGPT, it has been suggested that this development has the potential to present a more optimistic outlook on the future of digital platforms.

OpenAI asserts that by employing artificial intelligence (AI) instead of human moderators, GPT-4 can facilitate “significantly expedited iteration on policy modifications, thereby diminishing the timeframe from months to hours”. According to OpenAI, GPT-4 possesses the capability to comprehend both rules and subtleties inside extensive content policy documentation, and can promptly adjust to policy revisions, hence leading to enhanced consistency in labeling.

For example, the blog article illustrates how moderation teams could mark content to indicate whether it is inside or outside of the guidelines of a certain site. GPT-4 might then take a similar collection of data and label it differently, without understanding the answers ahead of time.

The moderators will be able to compare two labels of data. The moderators may then use any inconsistencies to clarify their rules and lessen the likelihood of confusion resulting from the application of the rules. In other words, GPT-4 might play the role of an average user in order to evaluate the rules and determine whether or not they make sense.

Currently, humans moderate content on a variety of websites. It regularly exposes them to potentially unlawful, violent, or hazardous information. We’ve seen time and time again what a terrible toll content moderation can have on individuals. Moreover, Facebook is paying $52 million to moderators who had PTSD as a result of the stress of their work.

All the AI models are getting trained for the mental stress that humans experience while dealing with difficult content. Reducing stress on the human moderators will assist to improve their working conditions. In such scenarios, humans will be able to be deployed without fear of burnout or PTSD.

The use of AI in this way prompts questions about the impact on employment. It is likely to worry that former content moderators will be rendered redundant. However, notwithstanding the fact that content moderation is not necessarily an enjoyable job.

In the same blog post by OpenAI, the individual’s call to the content platforms and social media websites is not addressed. However, there is a fear of loss for big companies to use AI as a cost-cutting technique without considering the consequences.

Still, if AI can reduce or eliminate the mental devastation faced by the overworked and underappreciated teams who moderate content on the websites used by billions of people every day, there could be some good in all this. It remains to be seen whether that will be tempered by equally devastating redundancies.

However, there is always a positive side to every technology that emerges. If AI can reduce mental stress and devastation faced by moderators of social content websites, then it is believed to be fruitful. But still, the time has to decide which scoop of AI will be offered to the moderators.

The real question is when will the AI Bill of Rights be drafted? What will the Bill of Rights include? If you know any answer to any of these questions, let us know in the comments section.

Check Out: OpenAI Introduces Customized Instructions for ChatGPT.