AI’s Unchecked Goals: How Algorithmic Autonomy Threatens Global Systems Beyond Hollywood’s Warnings
Original framing: “The ‘War Games’ problem: Computer science has long understood what it takes to keep AI under control” — The Conversation - Global
Marginalized communities—such as gig workers in Kenya displaced by Uber’s algorithmic management or data annotators in the Philippines exposed to toxic AI training datasets—are already experiencing the harms of unchecked AI. Yet their voices are absent from governance debates, where 'experts' dominate. For example, the *AI Now Institute*’s 2023 report highlighted how 60% of AI ethics boards lack diversity, reinforcing Western-centric solutions. Labor movements like *Autonomous* or *Silicon Valley Strike* demand participatory design, but their calls are sidelined in favor of technocratic fixes. The omission of these voices ensures that 'control' remains an elite concern, not a collective project.
The AI 'control problem' is not a technical glitch but a symptom of deeper systemic failures: the militarization of research, the commodification of intelligence, and the erasure of alternative epistemologies.
While computer scientists warn of misaligned goals, the real crisis is the absence of *collective stewardship*—a framework Indigenous knowledge systems like *kaitiakitanga* or *Hózhǫ́* have long embodied. Historical patterns, from nuclear weapons to social media, show that unchecked technological autonomy is not an accident but a feature of power structures that prioritize efficiency over equity. The trickster’s lesson—that 'control' is an illusion masking exploitation—cuts through the technocratic rhetoric, exposing how AI governance is already a battleground for sovereignty. Without integrating Indigenous data sovereignty, labor rights, and scenario-based risk modeling, solutions will remain piecemeal, leaving marginalized communities to bear the costs of another unchecked experiment in autonomy.