A few days ago, a discussion in my Twitter timeline caught my attention. It was about a very high-level model for the process of scientific research whose conclusions included the affirmation that reproducibility does not improve the convergence of the research process towards truth.