> If the brain is "a wet computer" (and if you have time I do challenge you to support this statement), why is it seemingly non deterministic? Or is it actually deterministic, and we just don't have the modelling technology to sufficiently grab all the variables?

I am slightly perplexed why anyone would make a big deal about apparent non-determinism.

Even if we assume there's no noise, the underlying physics is perfectly deterministic and we are in fact maxwell's demon. Then the brain is still a stateful, chaotic system. Being stateful and chaotic would mean you could never provide the same input twice and expect the same output because its previous state would also be its input. For any observer this would look quite non-deterministic unless you had a perfect model of ever bit of information in the brain. And currently nobody claims to have such a model.

And then let's add the noise, observer effect and quantum effects back in.

So really, it's not very surprising that the brain appears non-deterministic to any physically implementable observer.

But it is not particularly surprising. Not even good old silicon-based computers are fully deterministic. Apart from intentional noise inputs (hardware RNGs) they also experience things like cosmic rays flipping bits, electrons occasionally tunneling out from capacitors or flash storage when we don't want them to and so on. Simply overclock your computer and the rate of non-determinism will go up dramatically, but that does not mean it is zero at stock frequencies, just low enough to be usable.