Agentic AI Advancements
Updates on the latest way to connect science to runners through the power of Agentic AI
A training plan has always been a compromise with a constraint: a coach cannot read every run of every athlete, so they hand you a template and you adapt it yourself, badly. Find a twelve-week plan at ten weeks out and you skip the base. Find it at sixteen and you invent two weeks of your own. The plan cannot tell you which; it was frozen before it met you.
These constraints are going away, and almost nobody in this sport is writing about what happens next. This is where we work it out in the open: what a plan becomes when it can actually respond to you, what a model has to understand about physiology before it is allowed to prescribe anything, and what we get wrong building it.
