billions of tubulin in brain neurons, estimates of computational capacity of human brain

In The Age of Spiritual Machines from 1999, Ray Kurzweil used the estimate for the capacity of the human brain as 100 trillion connections, and 200 calculations per second per connection, 20 million billion calculations per second, 2 * 10^16.

His numbers are what led to the hype around an imminent technological singularity, and organizations like SingularityU. I believed in it for 2 years between 2012 and 2014. I still think there are lots of value in a lot of the work, technology does evolve at exponential rates, none of that is false, it just has to be re-framed in the context of microtubules as computers. He is using 200 calculations per second per connection, and 100 trillion connections, 200 * 100 * 10^12 (or “20 million billion”). Using just his numbers, and if its n-times higher, 2 * 10^16 * 10^n, 10000x higher is 2 * 10^16 * 10^4 = 2 * 10^20, and if doubling in capacity every year, 10000 doublings is log2(10000) = 13.3 years. With a billion times more calculations per second per connection, 2 * 10^25, log2(10^9) = 30 years.

Microtubules operate in the mega hertz or something like that, clock frequency, tubulin in the giga hertz, Stuart Hameroff says that they switch at 10^7 computations per second, and that neurons as a whole operate at 10^15 cps.

10^15 is 1000 trillion, if 1000 trillion more calculations per second, 200 * 10^15 * 100 * 10^12 = 2*10^31, log2(10^15) = 50 years. When I first discovered that microtubules were computers, 3 years ago, I was thinking it would push the singularity forwards with around a century. Overall it becomes a very different conversation.