I was disappointed when I first looked into the C. elegans emulation progress. Now I’m not so sure it’s a bad sign. It seems to me that at only 302 neurons the nervous system is probably far from the dominant system of the organism. Even with a perfect emulation of the neurons, it’s not clear to me if the resulting model would be meaningful in any way. You would need to model the whole organism, and that seems very hard.
Contrast that with a mammal, where the brain is sophisticated enough to do things independently of feedback from the body, and where we can see these larges scale neural patterns with scanners. If we uploaded a mouse brain, presumably we could get a rough idea that the emulation was working without ever hooking it up to a virtual body.
If you’re talking about Charlie Stross’s Lobsters, yes this was inspired by Henry Abarbanel’s work. He ran around the office going “They’re uploading lobsters in San Diego!”
You need more than just muscle cells to do a whole-body emulation here—c, elegans has roughly 1000 cells all told (varies depending on sex; hermaphrodites have somewhat fewer, males somewhat more).
I was disappointed when I first looked into the C. elegans emulation progress. Now I’m not so sure it’s a bad sign. It seems to me that at only 302 neurons the nervous system is probably far from the dominant system of the organism. Even with a perfect emulation of the neurons, it’s not clear to me if the resulting model would be meaningful in any way. You would need to model the whole organism, and that seems very hard.
Contrast that with a mammal, where the brain is sophisticated enough to do things independently of feedback from the body, and where we can see these larges scale neural patterns with scanners. If we uploaded a mouse brain, presumably we could get a rough idea that the emulation was working without ever hooking it up to a virtual body.
The lobster stomach ganglion, 30 neurons, but a ton of synapses, might be better for since its input and output are probably cleaner.
Modeling lobster stomach ganglion work is going on at Brandeis and what they are finding is important: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2913134&tool=pmcentrez&rendertype=abstract
Given the results they are finding, and building on their methods it is not inappropriate to start thinking one level up to c. elegans
Also because there’s fictional prior art?
Maybe there’s fictional prior art because the lobster stomack might be better.
If you’re talking about Charlie Stross’s Lobsters, yes this was inspired by Henry Abarbanel’s work. He ran around the office going “They’re uploading lobsters in San Diego!”
“You would need to model the whole organism, and that seems very hard.”
There are only ~100 muscle cells. People are trying to model the the brain-body combination, but that doesn’t sound unreasonably hard to me.
You need more than just muscle cells to do a whole-body emulation here—c, elegans has roughly 1000 cells all told (varies depending on sex; hermaphrodites have somewhat fewer, males somewhat more).