Commentary — This MIT News Office story on how brain circuitry findings could shape computer design. Neuroscientists just discovered that brain cells do not compute information as binary systems.
For the record:
- Greek philosophers already objected to the binary nature of formal logic — in that there are things that are neither true, nor false
- Every century since the Renaissance saw philosophers champion this viewpoint
- Kurt Gödel proved this viewpoint in 1931 — he proves that a consistent system will sprout things that are neither true or false
- Algirdas Greimas introduced an interesting additional insight when he pointing out that semantic-wise, ‘true’ and ‘not false’ are not strictly equivalent
- Lotfi Zadeh introduced fuzzy logic in 1965 — in fuzzy logic, things are somewhere in between true and false, with a ‘neither true or false’ in the middle
- Cognitive scientists have been talking about heuristics for years
Other than that, all is fine in the best of worlds. Neuroscientists just “discovered” that brain cells do not compute information as binary systems.
I think the real interesting point in the article is the part that went mostly unnoticed by news coverage:
By demonstrating the existence of tiny excitation/inhibition modules within brain cells, the work also addresses a huge question in neuroscience: What is the brain’s transistor, or fundamental processing unit?
But how huge a question is this?
- Computer scientists have been talking about neural networks for years, and reporting that very formal things emerge from these sometimes very fuzzy graphs, and vice versa
- Claude Shannon’s theorem states that you can go from analog to digital and back with no loss of information, which means that in the proper conditions it is perfectly legal to implement a fuzzy -1 to +1 with a binary system, and vice versa
Yep, that’s right. This means that the brain’s transistor question is an irrelevant question. Nevertheless, neuroscientists will ultimately need to “discover” this too.