Fred Reed describes problems with a purely mechanistic world view that also point to why recording training and behaviour cannot guarantee results:
“…According to this view, nothing happens, or can happen, that does not accord with physics.
“This approach, mechanistic and deterministic, works well as long as the observer is not taken into account. Astrophysics predicts with near exactitude the motions of planets. Solid-state physics describes accurately the behavior of electrons in microcircuits. In textbooks of biochemistry one reads of stereochemistry and charged groups and catalysis and so on that in fact describe what happens. It all works.
“Grave problems arise when you take the observer—the scientist, you, me—into consideration. The obvious first problem is that of consciousness. Your brain is a complex structure undergoing complex reactions, but all of these reactions follow the laws of physics. Yet nonetheless you are conscious. Is this something outside of physics? If so, then we have the sciences on one hand, and Something Else on the other, and the question becomes how they interact. Or is consciousness a physical variable, like gravitation? If I give you a large injection of Demerol, you will lose consciousness, and the biochemical mechanism can be given—but that doesn’t explain what consciousness is.
“Then there is the vexed question of volition. The end points of physical systems are determined by the starting conditions: The final positions of balls on a pool table depend entirely on the initial velocity of the cue ball, elasticity, coefficient of friction, and so on. The same determinism applies to chemistry: mix identical quantities of identical chemicals under identical conditions, and you get statistically identical results. If this weren’t true, chemical engineers would be in a helluva fix.
“So how can you choose to do one thing instead of another? Why is your “decision” not completely determined by the starting configuration of your brain? This is certainly true of computers which, given the same program and the same inputs, will always produce the same results.”
All Watched Over By Machines Of Loving Grace, Part 3 -The Monkey in the Machine and the Machine in the Monkey describes a mechanistic view of humans as machines. It is an error that is also at the heart of the UKAS ethos.