3 ms·
The problem is that you never know where better optima lie. You have to guess. It is easiest when a new optimum is easy to reach by just following the gradient
by nine_k 6y ago
The problem is that you never know where better optima lie. You have to guess.
It is easiest when a new optimum is easy to reach by just following the gradient. Short-term thinking works well here.
There may be some optima much farther down, which take a lot of "buried" investment to reach: efficient photovoltaics, electric cars, reusable space launch systems, maybe even fusion power. We know that some of them are better optima than the previous state, because they have been reached. Some, like fusion power, haven't been, but there is a lot of hope that they'll work. Some other attempts, like supersonic passenger aircraft, ended up not as optimal as their proponents had hoped. You never know.
The problem is that you can only reach these distant goals by concentrating resources and spending them without an apparent commercial result for years and decades. This takes a high concentration of resources and a commitment. Currently only billionaires and certain national governments can afford that. Most public companies are too beholden to their quarterly results to follow long-term policies which do not bring steady profits along the way, and many governments are in the same situation because the voters do not care about long-term goals enough.