Slashdot Mirror
Brain Prosthesis Ready For Testing
jhouserizer writes "New Scientist is reporting that an artificial hippocampus is ready to undergo testing. The leader of the team of scientists is Theodore Berger of the University of Southern California in Los Angeles. They hope these artificial hippocampuses can replace damaged (stroke, Alzheimer's, etc.) portions of your brain. I wonder what portions of 'you' would be noticeably different to your family & friends? I wonder how long it will be before we can have HUDs, such as in this story by Cory Doctorow?"
Anyway, when I first saw this headline I was thrilled at the prospect of scientists anywhere actually understanding any chunk of brain well enough to replace it (with a semiconductor, no less -- those must be some awesome I/O buffers on that ASIC -- what's brain voltage, uV? nV?).
Then I saw:
They just brute-forced it! This is remarkable achievement, but moreso from tech implementation standpoint than a brain understanding standpoint.
The point is, we don't have any clue at all about the uber-divx format that encodes human perception or memory. So the idea of storing it outside of the brain (or even viewing it, or cross-connecting 2 brains) is kinda silly at our level of understanding.
We just found a little chunk (the hippocampus) that is essential to storing memories and happens to get whacked often enough by stroke and such. Then we did an all-possible-input-combos test on this chunk (using rat brains, apparently), recorded the outputs, and burned the whole thing into a look-up table in a chip, and (this is the cool part) connected the chip to a real brain, bypassing a broken hippocampus chunk.
We just mimicked a relatively simple part of the brain with an exhaustive, brute-force approach that may not scale well to human hippocampi.
everything in moderation
I do research on hippocampal functioning --- more specifically, I build neural network models of how the hippocampus supports memory for specific events. I was surprised by the statement in the new scientist article that "we know nothing about how the hippocampus encodes memories". There is actually quite a lot of consensus among researchers as to how the different subregions of the hippocampus support its overall function of storing and retrieving memories. If you want more information, a good place to start might be this paper that I wrote with my colleague Randy O'Reilly. Go to:
e mo ry/
http://www.princeton.edu/~psych/PsychSite/compm
then click on the first article under "Review Papers". You can follow the references to find other, relevant papers. Also, I should say that I am extremely skeptical that the prosthesis described in the New Scientist article will be able to substitue for an actual hippocampus. One of the key properties of the hippocampus (and the brain more generally) is that it *changes* as a function of experience --- every time you store a new memory in the hippocampus, it changes the strengths of synapses, which in turn changes the input-output function. So I can't see how it would be possible to replace the hippocampus using a simple, static lookup table. I may be missing something, but I think we are still a very long way from building an artificial hippocampus, and I think that we won't be successful in this endeavor unless we build in some knowledge about how the structure actually works...