Amazing new computer accessory: it’s called a ‘mouse’
Virgil:
April 30th, 2007
Researchers from the US have managed to simulate half of the brain of a mouse (the squeaky cheesy kind) on their supercomputers. This is incredibly interesting in the context of the philosophy of mind and neuroscience (the latter being much more the purpose of the research). All brains, including those of humans, are made up of a vast network of interconnected neurons. Neurons act like messengers – they receive signals and pass new signals on. As it is very commonly believed that brains are responsible for thought, decisions, memory, rationality and consciousness, it seems logical to suppose that in some way these mental properties are the result of the spaghetti of signals in your neurons.
The wonderful thing about neurons is that they can be simulated. This is because they work exactly like a computer function, a piece of code that can receive and send on signals. This idea has been used for many years to create special programs called “neural networks” which are capable of some very limited brain-like learning for specific tasks. In a similar way, this new research has mapped half of the neurons from a mouse’s brain onto computer functions in a neural network called a “cortical simulator” – a simulated half-mouse.
When the program was run, the artifical cortex had “biologically consistent dynamical properties” – the “nerves” fired in the characteristic staggered, co-ordinated patterns seen in nature. Half a mouse was inside the computer. Although this research is still very much in its infancy, perhaps one day in the future we’ll be able to attach a different kind of mouse to our computers.
April 30th, 2007 at 12:13 pm
I’ve always thought most analytical philosophy of mind could basically be replaced (not in a Churchland way – I think his “completed neuroscience” concept betrays a lack of understanding about the very nature of science – science “completed” pah!) with actual neuroscientific research. As I am not an analytical philosopher (though I certainly like elements of it, some very much indeed), perhaps you can explain what unique its views add to the debate that are not possible from pure science?
April 30th, 2007 at 5:21 pm
It’s a matter of the philosophy of mind, Alex, as to whether it can be replaced by neuroscience. This brute fact renders it irreplacable without a deductive argument to prove such a replacement, and I highly doubt there is such a thing! On the contrary, there are plenty of arguments to prove the opposite; for instance, you may want to investigate a certain thought experiment called “Mill’s Mill” or “Mary’s Room” (titles vary).
April 30th, 2007 at 7:03 pm
Good olde Mary – I was actually aware of that one, and the stuff about qualia – I am very much on the side of the “she would learn something new” side.
May 6th, 2007 at 5:59 pm
[...] Petaflops? First we must explain what a flop is. Flops? First we must explain why people build supercomputers. People build supercomputers because there is some ridiculously complex problem they want to solve. Suppose you want to simulate the behaviour of neurons in a mouse’s brain. Because neurons each have up to 8 thousand synapses, each synapse requires a calculation every time it is used, there are up to 8 million neurons in a brain, they all work at the same time, and they update every moment, you can appreciate how many calculations a second are going to be required in the simulation. This is why mice brains are still more powerful than our best super-computer, BlueGene/L, which was built by IBM. [...]