> similar to how in capitalism products become obsoleted in favor of newer, better ones. this is idealistic and I think ignores a very important effect that is

> similar to how in capitalism products become obsoleted in favor of newer, better ones. this is idealistic and I think ignores a very important effect that is present in both consumer capitalism and in scientific research: the institutional interests of those on the supply-side tend to dominate those on the demand-side. In scientific research the supply-side would be research funding institutions such as universities, and the academic-industrial complex of publications. in consumer capitalism we see many examples of products getting worse over time (from the consumer point of view) because it serves the interest of the producer. the best example I can think of is home appliances. In the 1950s and 1960s they were engineered for durability and maintainability. In 2017 they are engineered to be cheap to produce and to be disposable. Washing machine repair service is basically a thing of the past, for example. Now you are expected to just buy a new one. Similar effects can be observed in a lot of products. what might this effect look like in scientific research? a lot of the "replication crisis" is related to this effect. there is pressure to write up a positive and interesting result, rather than a negative result. there is also pressure to write up theoretical research that attracts attention even if it doesn't yield experimental procedures to test the theory. is this going on with string theory? I don't know, but maybe. my point is just to call into question the notion that there is some kind of positive selection pressure that improves research over time. maybe there is, but I wouldn't count on it being strong enough to overcome institutional pressures.