So as usual, I am not sure what to make of an intriguing article, therefore I am sharing it with you (from the BBC Future “Will we ever” series): Will we ever… simulate the human brain?. For those who may not have the time to make the jump to the article it is basically this: the European Commission has granted One Billion Euros to the Human Brain Project (an aside: One Billion is always capitalized when paired with a major currency – it’s a rule I just made up. I call it the Caraballo Currency Capitalization Convention). As the “Human Brain Project” name implies, it aims to build a computer-based “simulated” human brain within 10 years.
Two main points here:
- One Billion Euros is a whole lot of money. I mean, it’s like A TRILLION DOLLARS (see the above mentioned Caraballo Convention). Okay, okay. Actually, it is more like $1.3 Billion Dollars. Still, that’s a whole lot of €s and/or $s.
- Isn’t this how every dystopian sci-fi comic book, novel, movie, video game and music video starts? Try to do something awesome and it goes horribly, horribly bad? Did Terminators 1, 2, 3 or 4 (are there more?) not get released in Europe? Apparently not, since the Brits launched their fourth – fourth! – Skynet satellite in December 2012. At least we know who to blame when serving our robot overlords.
Now there are lots of good, scientifically sound reasons for building such a simulated brain, such as being able to beat all of the other scientists at chess, to count cards at the local casino, and of course, to beat up on all of the mere mortals on Jeopardy (Watson is now fully engaged in the fight against cancer, btw). While those are certainly fine reasons, the main purpose behind “creating” a human brain is it would (theoretically) help us better understand how parts of the brain interact with one another. Of particular interest is how one part of the brain compensates (or doesn’t compensate) for another part, without having to “test” (read: injure or impair) the actual brains contained in the actual heads of actual people. This would, in fact, be awesome, and revolutionize the care and treatment of those with severe brain injury or mental illnesses, and possibly even impact the way we organize and share information.
The main problem is – we don’t really know how the brain works right now, so how do we build a model that simulates it? Apparently we’ve been trying for decades and have a few models that can do a few things, but nobody has had the audacity and/or the funding to build something as complex as the entire brain. Until now.
I am all for bold research, but here’s the analogy that came to mind for me: early attempts at heavier than air flight. Some very smart and ambitious folks spent a whole lot of time and money looking at birds and building contraptions that tried to mimic the structures and movements of our feathered friends (just as the researchers in this project are building their model by mimicking the structures and activities they “see” in the brain). As it turns out, successful heavier than air flight (for humans) has nothing to do with feathers or flapping wings. Why would we think a successful “model” of the brain will “look” anything like the ones in our heads?
Another reason for concern – two telling quotes from some recent research conducted at the University of Iowa’s Neurological Patient Registry:
… self-awareness corresponds to a brain process that cannot be localized to a single region of the brain … In all likelihood, self-awareness emerges from much more distributed interactions among networks of brain regions.
Followed up with this:
Clearly, neuroscience is only beginning to understand how the human brain can generate a phenomenon as complex as self-awareness.
In short – we aren’t sure how the brain becomes self-aware, but we’re going to go ahead and build one. Hmmm. So, what if it happens to become self-aware in the process? Maybe it won’t try to destroy us, but would it be ok to hurt it, or turn in off? Then what?
Finally, there is this – according to a top researcher (not associated with the HBP) the “race” to build a human brain is really about “who can get the most biological functions and animal-like behaviours” from their program. So, super-brains that behave like animals and make mistakes? Great. At least the big (real) brains at Cambridge have the funding to investigate how to mitigate the risk of robot uprising as well. Hopefully they finish before the brain-builders!
So in case you couldn’t tell – I am not a fan of trying to replicate the human brain. In my opinion, the defining characteristic of the human brain … is the human heart.
What do you think?
Digital Brain – dailygalaxy.com
Dancing Robots – PC Watch – Japan
Abnormal Brain – Young Frankenstein (A great Mel Brooks movie, 1974)
Phrenology Chart – http://en.wikipedia.org/wiki/Phrenology
Red Bull Flugtag – Sydney Telegraph