The Evolving Role of Likes in AI Development and Social Media Dynamics

The article explores how social media ‘likes’ have become a critical resource for training AI systems to mimic human decision-making, particularly through reinforcement learning from human feedback (RLHF). Max Levchin, co-founder of PayPal, emphasizes the value of Facebook’s vast ‘like’ data for improving AI alignment with human preferences. However, AI is already reshaping the purpose of likes, as platforms like Facebook and YouTube increasingly use AI to predict user preferences, potentially making the like button obsolete. AI’s role extends to content generation, with growing amounts of liked content being AI-produced, raising questions about authenticity and the original intent of likes to incentivize human creativity. Examples include AI-corrected performances, like Alicia Keys’ Super Bowl halftime show, and virtual influencers such as Aitana Lopez, a computer-generated persona with 310,000 Instagram followers. The FCC has begun regulating AI voice cloning due to scams, highlighting risks of deception. The article concludes by questioning the future of the ‘like economy’ as AI-driven creators and consumers interact in a landscape where distinguishing real from synthetic becomes challenging, necessitating transparency tools to verify authenticity.