News

June 20, 2010. - Randal A. Koene
Come see us at the first public workshop on Advancing Substrate Independent Minds (ASIM-2010), August 16-17, organized by carboncopies.org, satellite to the Singularity Summit 2010 in San Francisco.

January 11, 2010. - Randal A. Koene

One man's mission to create a living mind inside a machine
(http://www.dailymail.co.uk/sciencetech/article-1240410/The-real-Frankenstein-experiment-One-mans-mission-create-living-mind-inside-machine.html)

November 24, 2009. - Randal A. Koene

Re: Recent simulation advances by the team of Dharmendra modha (http://spectrum.ieee.org/computing/hardware/ibm-unveils-a-new-brain-simulator).

You can guess how many people have been approaching me with this news (IBM's brain simulator at the scale of a cat brain) in the past couple of weeks (among others, Robin Hanson, who asked me for this comment). Dharmendra Modha has a very interesting and ambitious plan (in the context of SyNAPSE) - though it is important to note that his official goal lies in the area of A.I. rather than transfer of mind functions or whole brain emulation (WBE).

This recent demonstration of computing power in simulations of biologically inspired neuronal networks is a good measure to indicate how far we have come and when it will be possible to emulate the necessary operations of a complete human brain. Given the storage capacity that was used in the simulation, at least some relevant information could be stored for each updatable synapse in the experiment. That makes this markedly different than the storageless simulations carried out by Izhikevich.

I sort of anticipated that there is sufficient drive in this direction, i.e. the representation/emulation part of what it takes to achieve a functioning whole brain emulation. That's why I tend to emphasize as Problems to solve: the DATA problem (what do we need to acquire) and the COLLECTION problem (how can we obtain that data properly). The COLLECTION problem seems to include the very technical problem that we need a new method of in-vivo recording, one that can operate both at large scale and high resolution. At the very least this is needed for the purpose of learning about characteristic cell function within large networks operating in the natural manner, even if you are satisfied with a post-mortem WBE scheme a-la Ken Hayworth.

As for the official quest to create powerful A.I., well it remains to be seen how much that particular experiment will contribute. I think that Robin Hanson and I both think that the hard problem of intelligence/sentience lies not in accumulating the necessary quantity of computing cells. Huge quantities of cells and the ability to easily grow arbitrarily more have been available to nature for hundreds of millions of years. That sentient humans appeared so recently seems to indicate that the crucial problem has more to do with how you arrange the cells and their connections and what you get them to do.

Dharmendra Modha is also doing work on that front of course, especially by attempting to obtain the human connectome in terms of major (and soon minor) connective pathways. He's definitely one to keep a close eye on.

[Update: Robin Hanson's Blog comment: http://www.overcomingbias.com/2009/11/bad-emulation-advance.html]
Comments