Human Brain Project Kicks-Off
So it comes to pass that mankind is about to embark on a massive attempt to model the human brain computationally. This is very exciting for me personally because, as my regular readers are aware, the study of artificial life has been a long standing hobby of mine which inadvertently resulted in the creation of gaming technologies like Direct3D.
“The Human Brain Project kicks off Oct. 7 at a conference in Switzerland. Over the next 10 years, about 80 science institutions and at least 20 government entities in Europe will figure out how to make that computer brain. The project will cost about $10 billion euros — or about $1.3B in US dollars.”
It’s huge, ambitious, exciting and all that good stuff, but I must confess to having some hugely mixed feelings about this endeavor. When I see these giant government price tags I can’t help thinking; Or you could just wait for 10 years for the technology to catch up to your ambitions and solve it for $200M sort of like Craig Venter did with the Human Genome Project. That’s a minor concern however compared to my deeper instinct that it’s doomed to horrible costly never ending failure by virtue of simply starting out wildly misguided.
This line kind of says it all.
“The research hinges on creating a super-powerful computer that’s 1,000 times faster than those in use today. If you’re keeping track, that’s an “exascale” supercomputer, one fast enough to model a nuclear explosion or the complex, planetwide forces that shape the climate. Just a few years ago, scientists started using “petascale” supercomputers like Blue Waters at the National Center for Supercomputing Applications (NCSA) in Illinois that went online last year.”
Let’s break these claims down a bit for rational consumption. The research doesn’t “hinge” on creating a big enough computer, it hinges on developing a DEEP insight into the nature of biological computation. Without new insight (which costs nothing) all they will learn is that they have a huge computer gathering dust that they don’t know how to use.
Yeah… about that planet wide climate modeling… these geniuses IMAGINED that they could do that in 1997 and completely failed to predict anything correctly about how our climate evolves, in fact they also failed to launch two rockets which blew up over the Antarctic Ocean in an effort to try to figure out WHY their climate models were failing. Oddly, they got a decisive answer for the hundreds of millions they spent on failed climate satellite launches… your models don’t work for the same reasons your rockets explode on launch…. YOU ARE IDIOTS with PhD’s. I thought we mathematicians and computer scientists had been pretty clear on the whole “We can’t model turbulent systems with any amount of computer power AND we can never know the initial state of a real-world system with sufficient precision to simulate it even if we could” If that were even REMOTELY feasible we’d be hearing about these researchers funding their projects with Lotto winnings or from tapping out the stock market… hasn’t happened yet and neither of these problems are even in the same ballpark of complexity as climate and brain modeling. I think one of the REQUIREMENTS for getting government funding to computationally model ANYTHING should be demonstrating that you can match your grant dollars with your online AI Poker winnings. Back in 1997 while the IPCC geniuses where consuming our tax dollars pretending to predict the weather my Mathematica AI projects were supporting me with income. If you’re not bright enough to model simple economic systems that can pay for themselves, should you be getting government support? Again, I point to Craig Venter, who was not only smart enough to sequence the Human Genome with a fraction of the resources of the multi-billion dollar Human Genome effort… he was able to fund his work privately! It all smacks of a giant academic boondoggle to justify massive research grants forever.
Another reason for my skepticism about all of this is that every approach they are discussing involves studying and modeling a brain… as though they are complete organisms that can be raised in terrariums. Don’t we already know that when we raise a brain in isolation from the real-world it just goes nuts… in fact if we take a working brain and isolate it for even a few hours it goes nuts. I’m pretty sure we don’t need to spend 10 years and 1.3 billion dollars also discovering that brains don’t work very well when you remove them from bodies… Since nobody has ever managed to actually keep a living brain in a jar and measure it’s function we don’t even know if the rest of the organism is NECESSARY to it working correctly… by the way… I’m going to venture WAY OUT on a speculative scientific limb and assert that the answer to that great unsolved scientific mystery is that the complete organism is VITAL to a brains correct operation and development. Crazy, I know, now they’ll also be calling me a Human Brain Project Denier… clearly there is a consensus of more PhD’s than I can argue with that this is a BRILLIANT venture fully justifying the vast investment everybody is so eager to make. At least the Human Genome Project had a completely measurable goal. These guys seem to be suggesting that 10 years from now they’ll have a giant disembodied virtual brain doing brilliant rational things without any intimate connection to the rest of reality. I guess the whole “landing humans on Mars” thing was a little too “tangible” for everybody’s tastes… too easy for the rest of us footing the bill to measure their progress. These scientists are wising up however, the well must have dried up after the whole Higgs Boson discovery thing so it occurred to them that it would be much brighter to dream up a giant costly scientific endeavor that has no tangible goal!
Is it just me or is there something pathetically humorous about spending 1.3 Billion dollars trying to use a delusional blob of thinking jello composed of 76% water that consumes roughly 100 Watts of power to analyze it’s own design by building a computer many thousands of times larger and more powerful? They must have gotten the idea from Douglas Adams. I think the computers answer will be: “You are now no closer to duplicating a brain by that approach than your TRS-80 is to designing an iPad… may I suggest using your PhD’s to wrap some fish?”
I know what Moulder on X-Files must have felt like… I WANT to believe… but every way I examine these ventures I end up finding a weather balloon or a guy throwing dinner plates in front of a camera instead of an actual UFO. This doesn’t smell like science it smells like make-work research for academics.
In a tragic related news story:
Yeah… I’m stricken… how can we let the Swiss surpass us in silly research spending! Where is our national pride! I can just see those smug Swiss researchers sitting around in their lab coats sipping their Swiss Miss hot chocolate and picking their teeth with those little Swiss Army Knife toothpicks laughing at our incompetence at rationalizing MASSIVE spending on science fiction research projects. F**KING SWISS! STUPID CONGRESS! I’m shattered… really…