Science
21 Jul 14
21 Jul 14
1 edit
I was reading an article on the BBC website which was quite interesting. The European Commission has put a billion Euros into the Human Brain Project, but there's some controversy. Essentially they allowed the project to allocate money internally themselves and internal politics seems to have meant that the bottom up approach, which seeks to model entire human brains based on the components, has squeezed out the top down approach, which takes a cognitive neuroscience approach. Much as I favour the bottom up approach I used to use in fundamental physics, I doubt it's the right way forward in neuroscience. I wonder if they should be forced to generalize their approach, or if there's an opportunity for countries outside the EU to poach some brains?
Edit: I forgot the link: http://www.bbc.co.uk/news/science-environment-28373264
21 Jul 14
Originally posted by DeepThoughtIs this doing it in software? I think in software they are up to 1/4 th of a cat brain, something like that.
I was reading an article on the BBC website which was quite interesting. The European Commission has put a billion Euros into the Human Brain Project, but there's some controversy. Essentially they allowed the project to allocate money internally themselves and internal politics seems to have meant that the bottom up approach, which seeks to model enti ...[text shortened]... some brains?
Edit: I forgot the link: http://www.bbc.co.uk/news/science-environment-28373264
22 Jul 14
Originally posted by sonhouseThis is the thing, they are spending all the money running a simulation of a human brain on a machine, rather than distributing the cash amongst different approaches. So yes it's a simulation, they aim to be able to do it in 10 years and need an exaflop speed computer.
Is this doing it in software? I think in software they are up to 1/4 th of a cat brain, something like that.
Here's another BBC article talking about the controversy:
http://www.bbc.co.uk/news/science-environment-28193790
and this is the URL for the project's website:
https://www.humanbrainproject.eu/en_GB/home
22 Jul 14
Originally posted by DeepThoughtIt seems to me reasonable to let the scientists in question make the decision. Their decision may be based more on politics than on rationality as you say, but an externally mandated decision is even more likely to be politically driven - given that the decision makers are less likely to be specialists in the field.
I wonder if they should be forced to generalize their approach, or if there's an opportunity for countries outside the EU to poach some brains?
I personally favor the bottom up approach.
22 Jul 14
Originally posted by DeepThoughtThey should be putting most, if not all, the money into parallel computing development. This would presumably serve both approaches, or is there something about the other approach that requires less computing power?
This is the thing, they are spending all the money running a simulation of a human brain on a machine, rather than distributing the cash amongst different approaches. So yes it's a simulation, they aim to be able to do it in 10 years and need an exaflop speed computer.
22 Jul 14
Originally posted by twhiteheadThey don't seem to have set up a transparent system of governance. Besides, there's such a thing as external audit and scrutiny - this is $1 billion of taxpayers money.
It seems to me reasonable to let the scientists in question make the decision. Their decision may be based more on politics than on rationality as you say, but an externally mandated decision is even more likely to be politically driven - given that the decision makers are less likely to be specialists in the field.
I personally favor the bottom up approach.
Regarding your other post, I think the two approaches should be regarded as being complementary rather than in rivalry [1] with each other. There's not much point setting up a bottom up thinking brain if we don't know what it's thinking about.
I think the terms of the project seem to not have been very clearly set out before they received funding. The US has a similar program and the funders seem to want to compete with that instead of taking a balanced approach.
[1] I'm stuck for a word here, contradiction maybe or exclusive to each other.
22 Jul 14
Originally posted by DeepThoughtI agree that the use of tax payers money should be transparent.
They don't seem to have set up a transparent system of governance. Besides, there's such a thing as external audit and scrutiny - this is $1 billion of taxpayers money.
Also, having read the article, I realised that many of those criticising it were experts in the field.
But without the transparency I cannot say much more about whether their decision was the correct one.
It does seem like an awful lot of money - but if it is going into developing parallel computing techniques then the side benefits may be significant.
If they took the other approach you mention, what would the money be spent on?
23 Jul 14
Originally posted by twhiteheadI don't really know what cognitive neuroscience is. I can make some guesses, but was just going by what it said in the article. At some point I'll read the Wikipedia page and might be able to form an answer. There's a good book called: "The Man Who Mistook his Wife for a Hat and other Clinical Tales" by Dr. Oliver Sachs, a neurologist. In it he describes the problems people faced because of neurological injuries. For example one woman had brain damage for some reason I can't remember and had lost the part of her brain which kept track of where her limbs are. A sense most of us don't realise we have - but you can close your eyes, reach out, and touch your finger tips together without peaking. She couldn't after the injury. She took it quite well, but had to be careful to watch what her hands were doing thereafter, because she didn't know where they were. I suspect that that is the general field that cognitive neuroscience lives in. I could very easily be wrong. Quite what they were hoping to get funding for in the Human Brain Project is not something I can answer.
I agree that the use of tax payers money should be transparent.
Also, having read the article, I realised that many of those criticising it were experts in the field.
But without the transparency I cannot say much more about whether their decision was the correct one.
It does seem like an awful lot of money - but if it is going into developing paralle ...[text shortened]... be significant.
If they took the other approach you mention, what would the money be spent on?
Originally posted by DeepThoughtA more cost effective way of doing this would be to find someone whose brain is already in working order and use that brain. This would keep the cost down and they wouldn't have to wait 10 years before using it.
This is the thing, they are spending all the money running a simulation of a human brain on a machine, rather than distributing the cash amongst different approaches. So yes it's a simulation, they aim to be able to do it in 10 years and need an exaflop speed computer.
Here's another BBC article talking about the controversy:
http://www.bbc.c ...[text shortened]... 0
and this is the URL for the project's website:
https://www.humanbrainproject.eu/en_GB/home