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 possible, that optimisation is itself nested inside arbitrariness.
3. What is optimal varies across environments/perspectives. Now you can call that a lack of concordance of interests, but it doesn’t require deliberate conflict, as implied in that paragraph. The best diet for my health is probably not the same as the best diet for a diabetic, nor is it necessarily the best diet for the environment, etc.
Then there is the issue of optimisation being nested in a social context, and in time. People who have already made choices will want to see the world in such a way that those choices were optimal. For example, doctors who perform circumcisions don’t want to believe circumcision is harmful, because they don’t want to see themselves as baby-mutilators. The optimal beliefs for them to hold to continue living their lives are not necessarily the same as the optimal beliefs for them to hold to be factually correct. This means that often when people talk about what is optimal, they’re actually optimising for a whole lot of past and context that isn’t visible to their audience, which for all intents and purposes to a listener are arbitrary.
That was more like four points, but oh well. In all, I think it’s less of a dichotomy than it seems at first glance, and the people who favour optimisation are either unaware or in denial of its infirmity.
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 possible, that optimisation is itself nested inside arbitrariness.
3. What is optimal varies across environments/perspectives. Now you can call that a lack of concordance of interests, but it doesn’t require deliberate conflict, as implied in that paragraph. The best diet for my health is probably not the same as the best diet for a diabetic, nor is it necessarily the best diet for the environment, etc.
Then there is the issue of optimisation being nested in a social context, and in time. People who have already made choices will want to see the world in such a way that those choices were optimal. For example, doctors who perform circumcisions don’t want to believe circumcision is harmful, because they don’t want to see themselves as baby-mutilators. The optimal beliefs for them to hold to continue living their lives are not necessarily the same as the optimal beliefs for them to hold to be factually correct. This means that often when people talk about what is optimal, they’re actually optimising for a whole lot of past and context that isn’t visible to their audience, which for all intents and purposes to a listener are arbitrary.
That was more like four points, but oh well. In all, I think it’s less of a dichotomy than it seems at first glance, and the people who favour optimisation are either unaware or in denial of its infirmity.