'Accelerating scientific progress through Bayesian adversarial collaboration' -- out now in the latest issue of Neuron. @hohwy.bsky.social, Karl Friston & I propose a formal framework for designing & evaluating adversarial tests of competing theories.

Out today: Our (post-Popperian) Bayesian approach to Adversarial Collaboration. Applies to all AC, from Eddington, over Kahneman, to multiexperiment ACs a la TWCF Cogitate. Check lead-author Andrew Corcoran's X-🧵 (+Free-print on SSRN) www.cell.com/neuron/fullt...