Deriving free will from singularities

William Tomos Edwards writes in Quillette to defend free will:

[Biologist Jerry] Coyne dismisses the relevance of quantum phenomena here. While it’s true that there is no conclusive evidence for non-trivial quantum effects in the brain, it is an area of ongoing research with promising avenues, and the observer effect heavily implies a connection. Coyne correctly points out that the fundamental randomness at the quantum level does not grant libertarian free will. Libertarian free will implies that humans produce output from a process that is neither random nor deterministic. What process could fit the bill?

No, they are both wrong. Libertarian free will certainly does imply that human produce output that is not predictable by others, and hence random. That is the definition of randomness.

Quantum randomness is not some other kind of randomness. There is only one kind of randomness.

Then Edwards goes off the rails:

Well, if the human decision-making process recruits one or more irremovable singularities, and achieves fundamentally unpredictable output from those, I would consider that a sufficient approximation to libertarian free will. Furthermore, a singularity could be a good approximation to an “agent.” Singularities do occur in nature, at the center of every black hole, and quite possibly at the beginning of the universe, and quantum phenomena leave plenty of room open for them. ...

The concept of a singularity becomes important once again here because if you can access some kind of instantaneous infinity and your options are fundamentally, non-trivially infinite, then it would seem you have escaped compatibilism and achieved a more profound freedom.

Now he is just trolling us. There are no singularities or infinities in nature. You can think of the center of a black hole that way, but it is not observable, so no one will ever know. There certainly aren't any black holes in your brain.

Coyne replies here, but quantum mechanics is out of his expertise.