News for hackers. Stuff that matters.
Re:Geoff Hinton Wants Computers to Think More Like Brains by criddell on Thursday January 03, 2019 @10:46AM (#18815925)
> The brain is solving a very different problem from most of our neural nets. You’ve got roughly 100 trillion synapses.

Hinton doesn't mention microtubules at all. Do any of the neural net or other connected neuron models incorporate the quantum vibration effects inside the neuron's microtubules?

Re:Geoff Hinton Wants Computers to Think More Like Brains by daveguy on Thursday January 03, 2019 @10:50AM (#18815968)
Not a neurophysiologist. Can you point me to any controlled studies verifying this microtubule quantum effects? If not, a principle based description of the theory and hypothesized effects would be helpful too. Microtubule physiology characterization would also be good. Completely not up to date on neurophysiology so pretty much any references would help. Thanks in advance!

Edit: Nevermind on microtubule resources. Plenty on the physiology references for those.

Re:Geoff Hinton Wants Computers to Think More Like Brains by criddell on Thursday January 03, 2019 @11:13AM (#18816158)
The quantum vibrations might be crackpottery, but the microtubules might still be information carriers[1]. I'm still curious if any neural net models incorporate these paths.

[1]: https://onlinelibrary.wiley.com/doi/pdf/10.1111/jnc.12621

Re:Geoff Hinton Wants Computers to Think More Like Brains by etiam on Thursday January 03, 2019 @04:45PM (#18819604)
Nets used for practical machine learning today are too abstract to incorporate them in any meaningful way. When you reduce the whole "neuron" to a weighted sum wrapped by some nonlinear function, in discrete timesteps, you've killed so much structure in the modelling it doesn't matter if the original information transfer path was supposed to be by membrane depolarization or by modifying cytoskeleton strands.

Arguably if there is a cytoskeleton mechanism for signalling or storage it's captured in some ways in changing the weights and evaluating the net, but not if it had any unique function beyond the abstraction which was already in place from the conventional view.

Studying other structures than the synapses and membranes is prudent, but Hameroff must be somewhere on the line between crackpot and intellectual poseur. Use your time on some real science instead.

Here's one option: http://www.gatsby.ucl.ac.uk/~lmate/biblio/dayanabbott.pdf

Re:Geoff Hinton Wants Computers to Think More Like Brains by aaaaaaaaaab on Thursday January 03, 2019 @10:57AM (#18816029)
I thought quantum microtubules are still considered crackpottery.
Re:Geoff Hinton Wants Computers to Think More Like Brains by lenticular on Thursday January 03, 2019 @01:06PM (#18817380)
There are definitely a few non-crackpots interested in quantum microtubules, but it seems very far from mainstream.

It's conceivable that quantum effects could matter in some way on very small scales, and that doesn't really have anything to do with quantum mysticism.

Re:Geoff Hinton Wants Computers to Think More Like Brains by codekilla on Thursday January 03, 2019 @12:19PM (#18816858)
Stuart Hammeroff gave a talk here at Caltech about them last year.....which surprised me a bit. It is still considered far afield from mainstream neuroscience, however, quantum biology is something many are seriously thinking about.
Re:Geoff Hinton Wants Computers to Think More Like Brains by criddell on Thursday January 03, 2019 @10:59AM (#18816047)
That could be. I'm the furthest thing from an expert in this.
Re:Geoff Hinton Wants Computers to Think More Like Brains by jostmey on Thursday January 03, 2019 @11:05AM (#18816101)
They are