Henry Markram, the director of the Blue Brain project, recently delivered a talk at TED that's gotten lots of press coverage. (It was the lead story on the BBC for a few hours...) Not surprisingly, all the coverage focused on the same stunningly ambitious claim, which is Markram's assertion that an artificial brain is "ten years away". I haven't heard the talk, so I don't know the context for the remark, but I did spend a few days with Markram a few years ago. The first thing to note about Markram is that he's incredibly brilliant and persuasive. I might be skeptical about the singularity, but I'd still be reluctant to bet against Henry and the Blue Brain team. The second thing to note is that the ten years timeframe - the one remark that gets spliced into every headline - isn't intended as a literal estimate. Two years ago, when I met with Markram, an artificial brain was "approximately ten years away." The point is that nobody knows how long it will take, since we don't even know what we don't know. I think Markram is using "ten years" as a way to suggest that the end goal - a simulation of human consciousness on a supercomputer - is both feasible and remote. It's technically possible, but only if the next decade goes according to plan. And when has that ever happened?
If you'd like to learn more about the Blue Brain, check out my old feature on the project:
In the basement of a university in Lausanne, Switzerland sit four black boxes, each about the size of a refrigerator, and filled with 2,000 IBM microchips stacked in repeating rows. Together they form the processing core of a machine that can handle 22.8 trillion operations per second. It contains no moving parts and is eerily silent. When the computer is turned on, the only thing you can hear is the continuous sigh of the massive air conditioner. This is Blue Brain.
The name of the supercomputer is literal: Each of its microchips has been programmed to act just like a real neuron in a real brain. The behavior of the computer replicates, with shocking precision, the cellular events unfolding inside a mind. "This is the first model of the brain that has been built from the bottom-up," says Henry Markram, a neuroscientist at Ecole Polytechnique FÃ©dÃ©rale de Lausanne (EPFL) and the director of the Blue Brain project. "There are lots of models out there, but this is the only one that is totally biologically accurate. We began with the most basic facts about the brain and just worked from there."
Before the Blue Brain project launched, Markram had likened it to the Human Genome Project, a comparison that some found ridiculous and others dismissed as mere self-promotion. When he launched the project in the summer of 2005, as a joint venture with IBM, there was still no shortage of skepticism. Scientists criticized the project as an expensive pipedream, a blatant waste of money and talent. Neuroscience didn't need a supercomputer, they argued; it needed more molecular biologists. Terry Sejnowski, an eminent computational neuroscientist at the Salk Institute, declared that Blue Brain was "bound to fail," for the mind remained too mysterious to model. But Markram's attitude was very different. "I wanted to model the brain because we didn't understand it," he says. "The best way to figure out how something works is to try to build it from scratch."
The Blue Brain project is now at a crucial juncture. The first phase of the project--"the feasibility phase"--is coming to a close. The skeptics, for the most part, have been proven wrong. It took less than two years for the Blue Brain supercomputer to accurately simulate a neocortical column, which is a tiny slice of brain containing approximately 10,000 neurons, with about 30 million synaptic connections between them. "The column has been built and it runs," Markram says. "Now we just have to scale it up." Blue Brain scientists are confident that, at some point in the next few years, they will be able to start simulating an entire brain. "If we build this brain right, it will do everything," Markram says. I ask him if that includes selfconsciousness: Is it really possible to put a ghost into a machine? "When I say everything, I mean everything," he says, and a mischievous smile spreads across his face.
My current project involves making a model of part of the very earliest vision-motor pathway in the brain. To nobody's surprise, we don't have the data yet to actually make a "real" model, with every neuron type accurately modeled; we still don't know exactly what neurons are involved in this circuit, in fact. So making a computational model of the entire brain in ten years is most likely just shorthand for "we're still very far away from our goal but please give us funding" and nothing else.
This really does remind me of the Human Genome Project. The original expectation was 100,000 genes. At the end the number was 35,000, then 26,000; latest estimate I've come across is 20,000. (Not to mention that there is a major philosophical argument about what actually constitutes a gene.) The implications of the declining numbers were and are huge. The 100,000 expectation was based and the same number of human proteins, and those proteins are still around, presenting an exquisite mystery for post-HGP researchers to work on (and on and on).
I hope this brain turns out to have a wonderful sense of humor, it'll need it.
The ten year time frame seems like a bit of a reach to me. Even Mr. Singularity himself Ray Kurzweil has this as a topic for debate 20 years from now.
I think that Markram's comment about exploring the philosophical issues experimentally is interesting. However, I don't agree with him in his belief that this will be sufficient to solve problems such as Nagel's. It all relies on the assumption that we can create a non-biological brain, which is the largest philosophical questions. If the answer to this question is negative, then the problem is undecidable from the standpoint of technology.
Also, the Blue Brain seems to ignore the possibility that the mind is dependent on the entire body (see Alva Noe's "Out of Our Heads"). Maybe the "brain" won't be entirely conscious until it is encapsulate in some robot form.
Modeling neuronal connections and activity is an admirable computational target. But let's not forget that we've barely scratched the surface of research figuring out neuronal-glia-astrocyte interactions work and how important they are in normal function.
The only thing we can be sure of is that the number of simulations needed to model a brain is way way higher than what's been proposed even here.
Central to the success of this experiment is acquiring enough power to run the model brain.
Henry Markram is very bold. He is also a brilliant neuroscientist. I think it will be interesting to see how his brain scales, how he factors in plasticity, and the evolving neural connections as learning occurs. Is it possible to build a "blue brain" containing a million processors (let alone 100 billion)?
We have set out at about the same time to consider building an artificial brain using custom electronic circuits containing nanotransistors. Our neurons have some plasticity, but building a "brain" with connectivities that evolve over time is not possible with current technologies. We believe a human brain to be 5 decades away, assuming our ability to manipulate nanodevices in real time is solved.
My biggest concern is that funding agencies will develop goals based on Henry's predictions, there will be disappointment and disillusionment, and it will be hard for smaller groups of serious researchers in the field to obtain funding after reality sets in. I have tried, when talking to the press, to be realistic about capabilities and timetable, and possibly err on the side of caution. I really hope Henry succeeds.
Haven't we already been through this (several times, over the past 60 or so years) with Artificial Intelligence? It is always 10 years away.
Hofstadter has some very apt comments as to how singularity is a long long way off...
What is the definition of an "artificial brain"? The definition implied by Makram is a machine that models a very limited subset of brain activities.
The human brain is the product of millions of years of evolution.
It is highly similar in many ways to, say, the house cat brain.
The very name of this blog - "frontal cortex" - reflects the fact that abstract and intellectual behaviors make up a rather small proportion of brain activity, and that much of human brain structure is highly analogous to that of other mammals, or even other vertebrates.
As we sleep, wake, breathe, eat, drink, go to the bathroom, establish social hierarchies, exhibit sexual behaviors, and so on, we tend to forget that the vast majority of human behavior is analogous to instinct-driven behavior seen in other animals.
Most of our motivations and emotional responses are directly related to our "older" brain structures.
Obviously, human "intelligence" is of great interest as well, to put it mildly. But intelligence serves our emotions and instincts, not the other way around.
The science fiction cliche of the computer or robot that "becomes so intelligent that it develops human emotions" is a silly human conceit. It has the process exactly backward.
I am interested in Ryan's comment (#4) where he speaks of Alva Noe's theories of consciousness as an activity that occurs in the interaction of brain, body and world. I know that Jonah posted his review of Noe's book "Out of Our Heads" this winter, but I'd really like to hear more about how this book of philosophy is received by the science community, and who in this community accepts and is taking action on Noe's idea that a fundamental change in the approach to brain research is necessary to developing AI such as the blue brain. Jonah - the ideas on both sides of the issue are quite complicated indeed, and I'd love to hear from the clearest cogsci blogger around!
If it even were possible to duplicate a human brain without a body, then, when you flick the on switch, it would immediately experience life with head cut off, all skin removed, no eyes, no nose no ears no proprioception. Not plausible.
Once they build the brain, how will the program it? Or will they just set off some random activations and see if it coalesces into something intelligent?
"an artifical brain is 10 years away" is code for "give me more funding and/or buy my book."
The 100,000 gene number was pretty much a handwaving guesstimate only used by some people. Those of us who worked in proteomics, before it had that name, knew that a single amino acid sequence could have a multitude of relative molecular weights and isoelectric points due to different post-translational modifications. Add to that alternative splicing of mRNA, and one gene might produce dozens of proteins. The 100,000 protein number never referred to 100,000 unique amino acid sequences; fewer genes is not a great mystery.
Blue Brain Computer Interface
Imagine your motor cortex fully activated while you have full muscle tone but both what your cortex says you are experiencing and what you are actually experiencing are not what you body is actually doing. You were trained to do this on a brain computer interface. Highly Skilled lucid dreamers in intense sessions and brain tomography on the level of seismic tomography make this all possible. Accessing the brain thru non-invasive means is vital in Berlin where Brain Computer Interfacers and the Locked-in are moving things with only their minds; however, one might say that all this research is treading water awaiting advances in Neuro-surgery. Iâm pitching the thoroughly developed non-invasive technique as a necessary prelude to the invasive interface. Iâm just looking for sympathetic places to post the story Iâm telling in the form of a fictitious photo journal.