[ExI] Vermis ex machina

Anders Sandberg anders at aleph.se
Sun Mar 1 21:36:24 UTC 2015

spike <spike66 at att.net> , 1/3/2015 3:33 PM:

From: extropy-chat [mailto:extropy-chat-bounces at lists.extropy.org] On Behalf Of Rafal Smigrodzki

>…Also, the brain is made of very low-reliability computational elements, with MTBF probably measured in hours (a guess on my part, not knowledge). Running a simulation using more reliable elements would allow for using a much smaller number of elements. What is the MTBF of a transistor?
Interesting point.  If a transistor is not run near its capacity, the MTBF of some configurations of transistor is still not known.  I have an HP calculator from college that still works.  I don’t know how many millions of transistors those things have, but every one of them is apparently operational still, which means the MTBF of that type of transistor used in that application would be many thousands of years, even using the most gloomy models we have.

Looking at a few data sheets like 
suggests MTBF on the order of 175 million - 6 billion hours at 25 C (the lower end for power transistors). That is, they can last hundreds of thousands of years. 

In the brain there are neurons that survive across the lifespan
so their MTBF is on the order of many decades. There are also neurons that recycle, and they have an annual turnover rate of 1.75%. So their MTBF due to turnover is about 57 years.

But this is cell failure. Synapses fail at proper transmission *nearly all the time*! 
Basically, there is a great deal of noise and variability introduced in synaptic transmission. The system is reliable since it uses many synapses and neurons, which are individually misbehaving a lot of the time. 

Anders Sandberg, Future of Humanity Institute Philosophy Faculty of Oxford University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20150301/735bd723/attachment.html>

More information about the extropy-chat mailing list