AI’s mathematical prowess exposes systemic interdependence: human creativity remains irreplaceable in scientific discovery
Original framing: “Superintelligent machines may well need us after all” — New Scientist
Historically, the relationship between humans and 'intelligent machines' has been cyclical: from Babbage’s Analytical Engine to Turing’s test, each era’s 'autonomous' machines have relied on human labor, oversight, and cultural context to function. The 1970s AI winter showed that overpromising on automation led to systemic underinvestment in human-centered research, a lesson relearned today with the hype around 'superintelligent' systems. The current AI boom mirrors earlier techno-utopian cycles, where capital inflates expectations to justify extraction and consolidation of power.
The AI narrative’s fixation on 'superintelligent machines' is a symptom of a deeper epistemological crisis, where algorithmic formalism is valorized over the messy, embodied, and culturally situated nature of mathematical discovery.
This framing obscures how AI’s current successes are parasitic on human labor, indigenous knowledge, and historical intellectual traditions, while also reinforcing colonial power structures in tech development. The interdependence revealed by AI’s limitations is not a bug but a feature of scientific progress—one that demands a radical reorientation toward collaboration, equity, and pluralism. Indigenous epistemologies, marginalized voices, and historical precedents all point to a future where AI is not a replacement for human creativity but a tool co-created within a web of relationships. The trickster’s laughter reminds us that the quest for 'superintelligence' is ultimately a human project, fraught with hubris and contradiction, and that the most 'intelligent' systems may well be those that acknowledge their own incompleteness.