Effort to Create Virtual Brain Begins

bryan8m writes "An IBM supercomputer running on 22.8 teraflops of processing power will be involved in an effort to create the first computer simulation of the entire human brain. From the article: 'The hope is that the virtual brain will help shed light on some aspects of human cognition, such as perception, memory and perhaps even consciousness.' It should also help us understand brain malfunctions and 'observe the electrical code our brains use to represent the world.'"

Comments

Thoughts on virtual thoughts

[IO ERROR]

All it takes to simulate a human brain is 22.8 teraflops? I thought I was smarter than that.

Seriously, they expect it to take a decade to complete. By 2015, we could probably get processors with that kind of power from the local computer store. Then everyone could have their own virtual brain...wait, are they going to GPL this?

So what happens if this thing develops a consciousness?

Re:Thoughts on virtual thoughts

[Bones3D_mac]

I'm more worried about how long it'll take for this thing to get bored, once it reaches that state. If they are going for the full human experience, how are they going to prevent sensory deprivation?

Will they use some kind of skin grafting onto a chip to let it "feel" things using the nerves in it, instead of simply simulating it with pressure/temperature sensors?

And what of other stuff like taste and smell?

Re:Thoughts on virtual thoughts

[smallfries]

Hold yer horses there laddie.

A neuron is *very* simple. Maybe just a sigmoid function over a sum. If thing actually is doing 22.8 terraflops (unlikely, I'm guessing that's the theoretical peak for the machine) then that gives 228 instructions per neuron. That is in the right range for operation.

There are not 'morphing' connections, they tend to mostly stabilize within the first few years of life. I can't remember the figure, its maybe on the order of a 1000 connections per neuron, so 228 floating point operations couldn't do this in real-time --- but who says it has to run realtime? Even at 1/10 speed this would useful.

I think the bigger problem isn't the processing power --- it's how you wire that network up. At the moment nobody has any idea how a brain is wired at the neural level. Imaging techniques just aren't accurate enough to tell us this. Getting your simulation to actually simulate a realistic brain would be an awesome challenge.

Obligatory...

[ArbiterOne]

"We marveled at our own magnificence as we gave birth- to A.I."

Obligatory HAL quote

[harlemjoe]

"without your space helmet Dave, you're going to find that rather difficult"

2001

Re:What if the simulated brain is a person?

[Vo0k]

A valid moral question.
Luckily the situation is more convenient. Call something like "suspend to disk", backup the whole state and you have the equivalent of hibernation. Can be "defrozen" and brought back to life anytime.

Brain != Thinking

[arstchnca]

For those who don't feel up to actually reading an article, the Blue Brain project does not intend to create artificial intelligence, but rather a replication of the physical side of the human mind - the brain. The 22.8 teraflops mentioned in the summary are going to be used to manage a database of "neural architecture." The whole project has little, if anything, to do with concsiousness.

As of this posting, there have been several "what if" posts about the project accidentally leading to the creation of artificial intelligence. Systems such as the fictitious Skynet will not rival the flexibility and depth of a single human mind until we fully understand the mind ourself. Lisa Fittipaldi, an astonishingly talented painter, is able to create beautiful scenes on what was once a blank canvas. At the same time, Ms. Fittipaldi is unable to paint an accurate portrait - she is blind.

We can only recreate what we understand.

Wishful thinking

[bloodredsun]

As someone who's spent many years as a neurophysiology researcher before becoming a programmer I feel I may have a bit more insight than the average person. What this project boils down to is a simplistic model of the simplist unit of operation of one area of the brain (neocortical column). Anyone who has followed research into areas such as epilepsy and memory will know of the massive gaps in our understanding of the realtionship of the brain and the mind. So this "first computer simulation of the entire human brain" is neither accurate in the sense that they are not simulating the human brain, nor are they the first to try what they are attempting. They only difference here is that they have the very public backing of a major corporation who understand the benefit of good publicity.

This sort of research is fascinating and despetately needs to be done, but it does no one any favours when people associate tabloid style headlines to it. The days when we wear Richard Morgan style "stacks" are still as far away as ever unfortunately.

Re:Wishful thinking

[markandrew]

my thoughts exactly; it may be a few years since i studied neural networks at university, but unless someone has sneakily made a quantum leap forward, any claims of simulating entire brains or creating self-aware computers is still science fiction.

it constantly amazes me that people still assume that once a certain amount of computing 'power' is available, a computer could suddenly become sentient, as if someone just flicked a switch.

we don't even know what sentience and consciousness really mean ourselves, and we've been trying to find out for centuries. as the parent poster said, we don't even understand everything about our own brains yet (and we're not that close, either), so claims of artificial life forms suddenly springing up after a few years of adding teraflops here and there are way off the mark.

we could create a machine which could count all the atoms in all the stars in the universe (every universe, if you believe in multiple ones) in a nanosecond, while simultaneously predicting every person's actions for the rest of time, but we'd still be no closer to a self-aware life form.

and even if we *did* create one, how could we tell if it was really 'alive', or just a brilliant simulation? the answer is, at the moment, that we wouldn't even be able to tell.

Re:Consciousness

[FirienFirien]

Zero evidence, yes. But we're looking to define the difference between a system that is alive (as we define ourselves to be) and one that is merely responsive. If you create a replica that works, at some point you cross the boundary that people define as unlife to life; and as logical and scientific we want to be about it, there will be those who consider it an aberration and will rise against it. The question will come of whether a soul is being made, whether it's founded or unfounded, evidence or not.

I've tried to keep the following part objective. It is not intended as a troll. Please read it objectively, and consider as part of a discussion over brain simulation and its repercussions rather than about religion. I believe what I believe, you believe what you believe.

I consider there to be no evidence - as such - for religion. Christians point to the bible, others to their own spiritual texts, but I'm quite cynical about the whole thing because there's no manifest evidence. But I don't go out and try to convince them that it's untrue, because I also don't have evidence to the contrary, and I'm also not fussed enough to feel an urge to bring people round to my way of thinking on that. However, as seen over and over (crusades, holy wars, jihads... the list goes on) the percieved insults against a religion are, often enough, responded to with force. US currency and (I think) the White House sigil bears the words "In God We Trust", even though the state is nominally unaffiliated with a religion; can you think of what would happen if it was motioned to be changed? Enough of the US population *believes* it enough that there would be outrage.

These same people believe that the creation of life, of soul, is for their God alone, and creation of new life by humans (other than the conventional way ;P ) would be blasphemous, arrogant against God, etc etc. This is the kind of thing that gets ranted about in churches. Whether or not there is "evidence" for it, to enough people it matters.