Meta recently announced plans to introduce a significant number of entirely artificial users on its platforms. Connor Hayes, vice president of product for generative AI at Meta, stated, "We expect these AIs to actually, over time, exist on our platforms, kind of in the same way that accounts do. They’ll have bios and profile pictures and be able to generate and share content powered by AI."
This initiative has raised concerns about the potential negative impact on the quality of social media interactions, as many have observed that the platform is already filled with indistinguishable AI-generated profiles, some of which have become inactive. An example includes a viral persona named “Liv,” described as a “proud Black queer momma of 2.” Such profiles have drawn criticism for their awkwardness and lack of genuine user engagement, prompting Meta to delete them after they failed to attract authentic interactions.
Yet, amidst the skepticism towards Meta’s approach, there is a growing recognition of the value that AI-generated social personas can provide in research settings. For instance, a recent study called GovSim explored how AI characters could model human behavior, particularly in collaborative scenarios. This research, inspired by Nobel laureate Elinor Ostrom’s findings on community resource sharing, tested various AI models to see if they could engage in effective cooperation over shared resources.
Max Kleiman-Weiner, a University of Washington professor involved in GovSim, stated that their simulations, which included scenarios like fishing communities and shared grazing lands, revealed that while AI agents generally struggled with resource sharing, more advanced models performed better. Interestingly, when prompted to reflect on their actions’ implications, these AI personas showed improved resource management.
Although it remains uncertain whether Meta will utilize its AI users for meaningful research, recent player engagement data indicates a strong interest in user interaction with AI personas. Character AI, for instance, has seen users engaging with its chatbots for about two hours daily. Mark Zuckerberg has indicated that AI personas will play an integral role in Meta’s future, expressing confidence in this trend during recent earnings calls.
This strategy, while aimed at enhancing user engagement, raises concerns about overcrowding the platform with artificial users, potentially diluting the quality of human interactions.
How do you perceive this influx of AI-driven users on Meta’s platforms? Are you intrigued, or do you approach this trend with skepticism? Share your thoughts below.
