I agree. My point being getting a larger sample size of things have diminishing returns, while being simple enough for most of the people to do it. With the end result of sizable expansion of the academia system without a corresponding increase in new advances. In fact, most of the things that are not theoretical in nature is already being spearheaded by the industry, likely for exactly that reason.
Given enough time the industry would accumulate institutional knowledge that creates a massive moat. The knowledge and experience is siloed up. Better have boring academia where everyone benefits than to continue feeding corporate titans.
