Why We Urgently Need to Invert the Human Stuck in the AI Loop

It was deeply disappointing to see Meta making moves in the race for 'superintelligence,' revealing their true priorities. Meta's new AI lab, pursuing superintelligence under Mark Zuckerberg, will notably include Scale AI founder Alexandr Wang as part of a major reorganization.

This partnership underscores a stark disregard by one of the most influential technology companies for the issue of human exploitation in the AI loop. While Meta aims for advanced AI, its foundation relies heavily on vast human-labeled data—a process largely facilitated by companies like Scale AI.

Scale AI's model uses a global workforce ("Taskers") for tedious, often psychologically demanding data annotation. This labor is essential for AI comprehension, yet investigative journalism and legal challenges reveal a troubling reality for many workers.

Numerous class-action lawsuits allege widespread wage theft and misclassification, denying fundamental employee rights like minimum wage and benefits. Investigations also highlight a profound psychological toll, as data labelers are exposed to disturbing content without adequate support, leading to trauma.

Critics like Karen Hao, in her book "Empire of AI" (Chapter 9: Disaster Capitalism), and Meredith Whittaker, President of the Signal Foundation and co-founder of the AI Now Institute, warn against this "scale at all costs" mentality. They argue it fosters "exploitation by design" and "AI colonialism," where the pursuit of "superintelligence" sacrifices human dignity and fair labor.

As we race towards an AI-driven future, we must acknowledge these realities. At this rate, not much of what it truly is to be human will be left in this AI loop. This is precisely why I've been advocating for inverting this dynamic and placing AI in the human loop. We must increase consumer awareness, surface these issues, and demand better legal guardrails. The exploitation of even one worker is merely the dawn of exploitation for all.

AI Now 2025 Landscape Report offers an alternate to current trajectory of AI and what it will take:

#AI #ResponsibleAI #DataLabeling #WorkerRights #AIintheHumanLoop #ScaleAI #MetaAI #Superintelligence #TechEthics #AIJustice #EthicalAI #DigitalLabor