Antifragility in ML: Taleb's Framework Applied

Taleb defines antifragile as systems that gain from disorder. Most ML regularization is just robust — it resists noise. We ran experiments specifically designed to test whether training can be genuinely antifragile.