Terrence Tao does not understand cancer medicine

This piece argues that Terrence Tao misunderstands how cancer medicine actually works. It contends that many widely used oncology drugs lack fully elucidated mechanisms - examples given include adriamycin, docetaxel, ibrutinib and venetoclax - and yet are adopted when clinical trials show benefit. Patients and clinicians routinely accept interventions with varying degrees of mechanistic plausibility and evidence strength: a positive phase 3 randomized trial would prompt near-universal uptake even if no one knew precisely how the therapy worked, and physicians sometimes act on promising phase 2 results, small case series, or preclinical signals when circumstances demand. The author acknowledges the danger of outright quackery but emphasizes the depth of desperation that drives acceptance of imperfectly understood treatments.

The deeper point is that mathematics and medicine operate under different epistemic regimes. Mathematical claims can be verified by formal proof checkers; human biology remains a vast, emergent system with many unknowns. Because of that empirical reality, clinicians prioritize robust trial evidence over complete mechanistic understanding and would welcome an AI-generated cancer cocktail that demonstrably improves outcomes in randomized trials. The piece insists mathematicians seeking solace from displacement by AI should not expect medical standards to mirror mathematical certainty.