Systemic shift: AI-driven autonomous labs accelerate scientific discovery, but who controls the agenda and at what cost?

Indigenous epistemologies view scientific discovery as a communal, relational process rather than an extractive, individualistic endeavor. The focus on autonomous labs reflects a Western mechanistic worldview that divorces knowledge from its social and ecological contexts. Traditional ecological knowledge (TEK) systems, such as those of the Amazonian tribes or Australian Aboriginal peoples, demonstrate how observation, intuition, and collective memory generate valid scientific insights without reliance on AI or automation. These perspectives are systematically excluded from narratives like Moore’s, which privilege institutionalized, capital-intensive research.

The push for autonomous science echoes historical patterns of technocratic control, from the Manhattan Project to the Green Revolution, where elite institutions dictated the terms of knowledge production. The DOE’s origins in nuclear weapons research reveal a continuity between military objectives and civilian science, with autonomy in labs serving as a new frontier for state and corporate power. The 19th-century eugenics movement, which also claimed objectivity, offers a cautionary parallel to today’s AI-driven research agendas. These precedents highlight how 'progress' narratives often mask extractive and oppressive structures.

Non-Western scientific traditions often prioritize cyclical, iterative knowledge systems over linear, predictive models. In China, the concept of *qi* and holistic medicine challenges the reductionist approach of autonomous labs, while African Ubuntu philosophy emphasizes communal benefit over individual achievement. The Global South’s reliance on open-source and grassroots innovation contrasts with the privatized, patent-driven model of AI science. These cross-cultural perspectives reveal how 'autonomy' in science is not a neutral term but a culturally embedded construct.

Autonomous labs leverage AI to accelerate hypothesis testing and material discovery, but their outputs are constrained by the biases of training data and algorithmic frameworks. Peer review and reproducibility crises in science are exacerbated by automation, as black-box systems obscure methodological flaws. The DOE’s focus on quantum materials aligns with military and energy interests, raising questions about the prioritization of applications over fundamental understanding. While AI can democratize access to data, it also risks entrenching epistemic hierarchies by favoring institutionalized knowledge.

Artistic and spiritual traditions often frame knowledge as a living, evolving process rather than a static product. Indigenous creation stories, such as those of the Navajo or Māori, depict science as a dialogue with the natural world, not a conquest over it. The surrealist movement’s embrace of chance and intuition contrasts with the deterministic logic of AI-driven research. These perspectives highlight the limitations of autonomous systems in capturing the ineffable, qualitative dimensions of discovery, such as beauty or ethical resonance.

If unchecked, autonomous science could lead to a bifurcation where elite institutions monopolize knowledge, while marginalized communities face further epistemic exclusion. Scenario modelling suggests that AI-driven discovery may prioritize commercially viable innovations (e.g., energy storage, weapons) over societal needs like healthcare or climate resilience. The militarization of research agendas risks normalizing surveillance and control in scientific practice. However, decentralized, community-led AI models could democratize knowledge, but this requires dismantling current power structures.

Women, people of color, and Global South researchers are underrepresented in AI-driven science, despite contributing to alternative knowledge systems. The narrative around autonomous labs centers white, male, institutional voices like Moore’s, erasing contributions from non-Western scientists or grassroots innovators. Indigenous scientists and activists, such as those resisting uranium mining on Native lands, are systematically excluded from these discussions. Their exclusion reinforces a cycle where the benefits of AI science accrue to the already powerful, while its harms disproportionately affect marginalized communities.