There are three main issues I have here with optimising.
1. It's not simple except in abstract situations, and even then only in some. Optimising the estimate of a gradient of a curve is one thing, but even something 'simple' in the real world like optimising the profits of a company is difficult beyond computability. Optimising something like useful life expectancy is even more absurdly difficult, especially if time spent on optimisation is deducted from the total.
2. What you optimise is, ultimately, arbitrary. Any 6 year old can prove this, with the 'why' game, and eventually the only answer is 'because I say so'. So while optimisation of a given thing may be... (read more)
There are three main issues I have here with optimising.
1. It's not simple except in abstract situations, and even then only in some. Optimising the estimate of a gradient of a curve is one thing, but even something 'simple' in the real world like optimising the profits of a company is difficult beyond computability. Optimising something like useful life expectancy is even more absurdly difficult, especially if time spent on optimisation is deducted from the total.
2. What you optimise is, ultimately, arbitrary. Any 6 year old can prove this, with the 'why' game, and eventually the only answer is 'because I say so'. So while optimisation of a given thing may be... (read more)