As simple as that thing may be...
...it's still smarter than Trump.
Boffins from the Technische Universität Wien in Austria have created a simulated brain, run it in a computer, and taught it a trick. "Brain" may be overstating things just a tad, because the simulation was of nematode C. elegans, a tiny worm that has just 300 or so neurons. We know that number because nematode C. elegans has …
Probably not. C. Elegans may only have 302 neurons, but a neuron is far more than a binary switch, and does a lot of 'internal processing' itself. They encode both digital and analog information, and operate as a complex machine in their own right. Though there is no sinple comparison between digital computers and the 'wetware' of a brain, it would make more sense to say that C.Elegans has a brain comprising 302 parallel processors.....
"Though there is no sinple comparison between digital computers and the 'wetware' of a brain, it would make more sense to say that C.Elegans has a brain comprising 302 parallel processors....."
Shhh! Don't tell Intel! They’ve got enough worms of their own already.
Indeed. I'm not sure if this is the same project, but there has been a recent EU project to model the nervous system of si.Elegans using several hundred FPGAs (pretty much the largest ones that were available at the time).
If I remember right the hardware is located at Nottingham Trent Uni.
I'm OK with that too so long as it is not these iWorms:
"14 cattle eyeworms removed from Oregon woman’s eye"
I for one would love to see the complete code. Is it like Game of Life? Write the code with a few simple rules then just sit back and wait for a pattern to emerge and then say, "look it did this".
I'm not belittling it, I love watching Mr Conway's creation, wrote my own version not long ago, just wondering exactly what these scientists have done.
I did find https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnx3d25pcDIwMTd8Z3g6NDQ3YjZhZTZiYWJiNDI5NA
but it's only a tiny algorithm. I guess they are working within a frame work and only applying algorithms within that framework.
cheers for that. Interesting. I downloaded the "code" but it appears to be data and code for manipulating the framework.
I was expecting a source code such as a complete "c" program or Python whereby the program lines went something like
IF nerveA TRIGGERS nerveB THEN
move muscle C
...but it turned out a little more complicated :)
I'll have to dig a bit deeper when I have more beer.
"Was there some kind of 'punishment' applied if the pole fell over and a 'reward' if the pole was balanced for a certain time?"
Then you've just given it the concept of 'work'!
Now all they need to do is train it to do this 'work' for the benefit of the 1%, and the worm would be a welcome addition to the rich and their 'race to the bottom' on cost.
Now map a leech - 10,000 neurons.
The mapping is a lot harder than the simulation. The c.elegans scan isn't taken from an individual - the species is unusual in having the position and connection of every neuron statically defined genetically, so they are all the same. It took a great many delicate dissections, each one uncovering a little more, to create a composite map.
No, the dissection is actually finer than that.
Take one C. elegans worm, fix it and embed it in suitable sectioning material. Then use a microtome to cut it into very, very thin salami slices and image each of these under a microscope. Finally, using a computer running much the same software as is used on CT scanners, reconstruct the animal's nervous system in 3D.
Repeat on a few different worms and there you have the nervous system of the brain (actually a nerve ring around the pharynx) of a nematode. Sit back and enjoy.
"Repeat on a few different worms and there you have the nervous system of the brain (actually a nerve ring around the pharynx) of a nematode. Sit back and enjoy."
Now for some applied research:
1. Repeat the investigation on the nerve ring around the anus.
2. Sell research to Essex Police.
It isn't just the brain, we can trace the origin of every single cell in the worm all the way back to a fertilised egg.
The problem is other creatures are not so determined in their cell lineages. Vertebrate development is rather fuzzier and rather more plastic than that in worms. But it is a nice thing to have done nevertheless.
The worm people are a bit humourless though. All the genes are alphanumerically named according to a schema. There is no sonic hedgehog gene in the worm, except there is and I can't be bothered to look up what they call it. It won't be something memorable though. Fun was had in taking single fruit fly genes named after mutation phenotypes and naming multiple vertebrate genes. Sonic was named because the natural world is not replete with different sorts of hedgehog, we had desert hedgehog and nobody complained about tiggywinkle hedgehog.
My favourites though are the fringes, radical and lunatic. A knowledge of Cold War political terminology is required to get it.
I'm confused. Where does the pole (on the tail) come from? Now I follow that they programmed the brain of the worm and, I presume, they also thus programmed something to mimic the 'body' (?) motion (to show output from the 'brain?)', but what about this pole?
Did someone leave a 'pole' subroutine lying somewhere in a library and the 'body' coder just accidentally tacked it on the end? Did they then magically notice the pole-code doing something? Or did someone randomly think "I know, I'll stick a pole on it's arse" - and if so, why a pole and not a beeper, or a LED, or a wifi link to a nuclear trigger?! How did they think "this will end up with it balancing a pole" if they didn't know the learning outcome?
(Even reading about this stuff is obviously waaaaay beyond me! Especially on a Monday. And a Friday. And all other days).
Biting the hand that feeds IT © 1998–2020