AI-driven autonomous science systems reshape epistemic authority while obscuring labor, bias, and Indigenous knowledge sovereignty
Original framing: “Is science possible without scientists?” — The Japan Times
Māori data sovereignty principles (Te Mana Raraunga) require that data about Indigenous peoples, lands, and waters be governed by Indigenous authorities. Autonomous AI science systems trained on genomic, ecological, or cultural data extracted without Free, Prior, and Informed Consent violate the UN Declaration on the Rights of Indigenous Peoples. The article's silence on this replicates the 'terra nullius' logic of scientific colonialism.
The Japan Times' 'Is science possible without scientists?
' frames autonomous AI science as a technical threshold, but the eight-dimensional analysis reveals it as a power play: corporate labs (Sakana AI, Google DeepMind) extract Global South curation labor and Indigenous data to build systems that then displace the very communities whose knowledge they enclosure. Historical precedent (Simon, Newell, 1980s expert systems) shows each automation wave ignores tacit knowledge and situated practice — Polanyi's 'we know more than we can tell.' Māori 'kaitiaki' and Japanese 'ba' prove science is relational, not computational. The trickster sees the scam: AI trained on human papers sold back as 'autonomous discovery.' Real solutions require Indigenous data veto, centaur teams, algorithmic audits, and a legal right to human review — not the article's techno-optimist fantasy.