I believe many software tools work like a vice for beurocrats making the beurocracy way less efficient.

Like, forcing some not very edge case friendly set of text field validation rules onto the beurocrats such that he can't just do what he tries to do as he could with paper work.

In many orgs. nowadays the processes seem to be made to fit the computer programs not the needs of the beurocracy.

And even if they produce "real" products. How many of them are needed. How many of them are redoing the same thing. And how many could in perfect world be done more efficiently by resusing some common solution.

Jobs funded by speculative investment that are highly sensitive to monetary policy are pretty orthogonal to the kind of thing addressed by "bullshit jobs" theory. They may be societally harmful or low-social-value more often than other jobs, but generally for very different reasons than bullshit jobs. They are "lottery ticket" jobs, where many are wasted, some are very valuable, and its hard to predict in advance which are which (for investor value; for various reasons endemic to capitalism more generally, this is a very imperfect proxy for social value, but that problem is also mostly orthogonal to the kind off harmful job "bullshit jobs" addresses.)