Francesco Zen

The interesting part of coaching was never writing the programme. Any competent coach can write a good programme. The part that takes years is knowing what to do when the programme meets someone’s actual week -- the flight that moved, the hotel gym with two dumbbells and a bench, the client who slept four hours and still turned up, the plateau that isn’t a plateau because they have been under-eating for a month and haven’t mentioned it.

Those judgement calls are what clients were really paying for, and they are the part that never scaled. One coach has a finite number of hours, and the people who most needed the judgement -- executives running sixty-hour weeks across three time zones -- were the hardest ones to give those hours to.

Mr Z exists because that specific kind of judgement turned out to be encodable. Not the motivational language, not the hand-holding; the decision rules. What to cut when someone has thirty minutes instead of sixty. When a missed week needs the plan rebuilt and when it needs leaving alone. What a set that looked wrong on camera actually means.

Mr Z is not a general-purpose chatbot with a fitness prompt in front of it. Its workout progression, its adaptation framework and the logic that decides what your next session should look like are modelled directly on how those decisions were made in practice, for real clients, since 2007.

That is the whole difference, and it is worth being plain about what it does and does not mean. It does not mean the AI is a clinician, and nothing it produces is medical advice. It means the defaults are a coach’s defaults rather than an average of the internet’s.