Saturday, January 13, 2007


In San Diego, researchers are uploading lobsters into cyberspace, starting with the stomatogastric ganglion, one neuron at a time. ("Lobsters" - Charlie Stross)
The lobster nervous system is very simple compared to mammals. The neural circuits that control the lobster digestive system, including the stomatogastric ganglion, have been extensively mapped and studied, and those neurons have, in a sense, already been "uploaded" as a part of computer models of the stomatograstric neural network. Indeed, Stross was inspired by this research. Popular Science editor Gregory Mone investigated further for his article science fiction "posthumans":
A few days after I return to New York from the Plokta conference, I find the San Diego researchers* on the Web and check with Stross to make sure they’re the right ones. Then I forward a link to the first story in Accelerando, the aptly titled “Lobsters,” to the scientists. A few hours later, a physicist in the group, Henry Abarbanel, calls me. He’s excited but a little confused. Excited that his team’s work helped to inspire a massive SF novel, perplexed because he can’t find any specific reference to their research in the story, although there is lots of stuff about uploaded lobsters. We talk a bit about science fiction in general—he was an Asimov fan as a kid—and then Abarbanel explains what he and his colleagues are doing with those lobsters.

The research, led by biologist Allen Selverston, focused on the California spiny lobster because only 14 neurons govern a key part of its gastric tract. This number of neurons is unusually small, which makes the area easier to model. Still, understanding the neurobiology of those 14 neurons was not easy. It took Selverston 25 years. Then Abarbanel and his colleagues needed two more to figure out how to re-create the system electronically. This work, too, was difficult: Abarbanel likens the process to having all the parts of a 747 laid out on the floor of a hangar with no instruction manual on how to put them together to make an airplane.

All that work, and they’ve electronically simulated just 14 neurons. That’s a far cry from uploading the 1011 neurons that make up the human brain. Naturally, I assume Abarbanel will laugh at the idea that uploading a human mind could ever be possible. But it turns out that he approves of Stross’s leaps of imagination. “Frankly, I don’t consider it to be crazy,” Abarbanel says. “Whether it’s five years or 10 years or 500 years, I have no doubt that we’ll figure out how to do it.”
Read "Lobsters", for Stross's take on what happens when the nervous systems of many lobsters are uploaded into the same computer.

*Abarbanel and Selverston are faculty at the UCSD Institute for Nonlinear Science . Selverston is the author of the stomatogastric ganglion article linked above.
Image from the National Oceanic and Atmospheric Administration


  1. Yes but if all my neurones are mapped, is that me? is that 1:1 map, is it the territory?

    And will the map think its me, or only think that it thinks that it's me?

    Thanks for this, I've been thinking about linking from short stories (in html or PDFs) to the science that is briefly mentioned in passing in them, and its good to seen it already done by others. I loved accelerando!

  2. I think those are good questions. So far we haven't been able to create an artificial intelligence that "thinks" like a person, let alone has self awareness. Can our sense of "self" be reproduced by simply "uploading" electronic equivalents of our neurons with all the same connections? I suspect no, but it would be cool if I'm wrong.


I've turned on comment moderation on posts older than 30 days. Your (non-spammy) comment should appear when I've had a chance to review it.

Note: Links to are affiliate links. As an Amazon Associate I earn from qualifying purchases.