I think it is going to be a long time before a robot is able to do my job. I could of course be wrong, but I hope not, and and the more I think about it, I don’t think I am.

A coaching robot would have to be an extremely comprehensive device. It would have to have extraordinary sensory input but also operate with a framework that properly understands the variety of ways that humans communicate and therefore learn.

“I’m sorry Dave, I’m afraid I can’t do that.”

It would have to understand the subtly perceived, but very complex, cultural aspects of people’s backgrounds. It would need to understand when visual or auditory cues are needed (not that there is some sort of learning style/coaching style matching that we want to get caught up it. Discussion for another time), what metaphors might be useful, what to emphasize and what specifically not to emphasize, or else create a new problem. This is how people learn, they mimic things, they take on corrections, they self-analyze, they improve.

Real time coaching in our life-time may never be super-ceded. Surely, it would be cost prohibitive to the extent that no robot anytime soon could become cheap enough that it could compete with a human coach.

It would have to anticipate and be ready to see the over-correction.

The over-correction is essentially what you at least want to see in real time. New athletes cannot go on feel, they do not have a ‘feel’. They will start at a position, move to the next position and move inefficiently. They will attempt to correct it based on feedback and will get that wrong. This is okay, this is necessary, and during this period, adaptive coaching strategies may find a quicker easier path to achieve the goal that departs from the starting model.



“You bubble-headed-boobie!”


The only frustrating reps in coaching are when you see no change. No use of control by the athlete upon their own body. The likelihood is that they behave as if they have bodies as opposed to that they are bodies. Skill installation and nervous system development is not the same as more common energy system stress. The exercising community is primarily used to perceiving energy system stress; much less so strength fatigue. They jog, they swim, they ‘condition’ they have laborious jobs they sense endurance failures. Strength is not a consideration because strength is never developed in any way close to their genetic potential and it is rarely exposed to its end range.

Generally the average person only considers strength when it comes time to help their neighbor move house. They get on the end of the couch and enter an unfamiliar world. This is more than just the energy system stress that their body more regularly encounters. Its not unusual for someone to live their whole life and develop very little of what their bodies are capable of physically, such is the limitation of what they develop. ‘Getting puffed’, ‘feeling tired’, ‘nausea’ is where their physical stress have taken them, not ‘how much weight? how far did it move?

Moving between static positions, developing locomotion, progressing to tonic or ballistic emphasis is something we have spoken about before. Its a basic conceptual teaching progression for physical skills. But a world where robots and computers program humans rather than the other way around would need to greatly advance.



Instead of assembling sums of data and plotting towards a pre-constructed end goal, there would need to be the ultimate role reversal. But whilst HAL 9000 and anything like ‘him’ is Heuristic and therefore uses algorithms to create rational decision making, adaptive humans working with adaptive coaches pose a much greater challenge than just a game of chess.



CAPTCHA technology is holding back the tide on spambots; meanwhile, human coaches are waiting to see the over-correction.