The article misses that we do not know how classical world that we perceive arises. Ultimately according to QM the whole universe is just single wave function with no probabilities. But this contradicts our experience of the classical world including that we measure probabilities.

So we need to explain that to avoid various very ad-hock notions of classical measurement device and similar routes.

There are attempts to explain the classical world using a notion of decoherence, but that have problems with little progress in the last, say, 20 years to address them.

Then there are ideas based on Everett works that perhaps it is a conscious itself that perceives the classical world and the real world is the wave function. But then such ideas do not explain the exact numerical probabilities.

The interesting resent suggestion was that the probabilities were not really physical but rather represent the lack of information in Bayesian sense.

In past that was used to construct various hidden variable theories, but the new take is that the classical notion that the state of a system can be fully described by its properties at some moment in time is wrong. One also needs to know some information about future to fully describe evolution of the system. This is very speculative, but at least it is fully compatible with relativity and explains probabilities. They simply represent lack of knowledge of boundary conditions in the future.