AI and wiring of the brain (Introduction)

by xeno6696 @, Sonoran Desert, Thursday, February 10, 2011, 00:18 (5015 days ago) @ David Turell

Developing AI equal to our brain is very unlikely. The brain's development is very plastic, and depends on stimuli present (Mom reading or speaking to baby) and mistakes made, which are then corrected. As a result trying to make an brain-authentic AI colmputer program based on brain wiring is very problematic:
> 
> http://www.physorg.com/news/2011-02-trial-error-brain.html-Again because the Von Neumann model of computation is completely wrong for emulating the human brain. -A neuron is simultaneously a CPU, memory, and I/O (input/output). And each neuron is connected to many, many others. -Von Neumann architecture completely separates these things into completely discrete units that share communications channels. After spending the last year studying computer hardware I'm fully convinced that true AI isn't possible until you can model brain connections the right way. -(In case you guys didn't know I've been embedding links in the text... it just looks more natural that way.) -The major breakthroughs in modern AI research has actually been in building many--really stupid--processor elements that work together to solve problems. -The main research project I have this semester is in working on some graph theoretic research using NVIDIA GPUs. -I can process 1 task across 448 simultaneous computations here; this is just a smattering of what the brain can do.

--
\"Why is it, Master, that ascetics fight with ascetics?\"

\"It is, brahmin, because of attachment to views, adherence to views, fixation on views, addiction to views, obsession with views, holding firmly to views that ascetics fight with ascetics.\"


Complete thread:

 RSS Feed of thread

powered by my little forum