[p1esk]

Given how robust a brain is against noise, I'd be surprised if any brain signals are more precise than an equivalent of 3-4 bits.

[jbay808]

I agree, and I think in practice the brain's noise floor is also much higher than the theoretical thermal-noise minimum. But I guess the main point is that once we acknowledge that even 32 bits is more than enough, the difference between an analog and digital machine loses a lot of its philosophical weight.