The BRAIN Initiative: BAM or BUST?
April 9, 2013
President Barack Obama is introduced by Dr. Francis Collins, Director, National Institutes of Health, at the BRAIN Initiative event in the East Room of the White House, April 2, 2013 (credit: Official White House Photo by Chuck Kennedy)
What is the BRAIN* Project about? What are its goals?
“Well, nobody knows, actually. I certainly don’t know. But it appears that no one else knows either.” So says Scicurious, a PhD in Physiology and currently a postdoc in biomedical research, on her The Scicurious Brain blog on Scientific American.
“Basically, BRAIN is a very fancy initiative, with a fancy name … and so far, no goals,” she says. “When they first announced the BRAIN project, there was some mention of mapping out all the connections in the human brain. That’s around 100 TRILLION connections made by 85 billion neurons. It’s a Herculean task at best.
“But the brain connections are more than individual, they are always CHANGING. Everything changes your brain. Whether you’re stressed, what languages you speak, whether you ate breakfast. All of these things, even the smallest, can change small connections in your brain and alter your future responses and behavior. So a map of all the connections in a healthy human brain (probably an adult male of around 35) from which we can determine who deviates from the norm and what that means? It’s an idea that is doomed from the beginning.
Furthermore, “the timeline on the BRAIN initiative is currently 10 years … really not that long. … We don’t know what the goals are, but I think the ‘dream team’ better be very careful coming up with them, or in 10 years we’ll we looking at a pile of empty promises.”
Is this really going to be the great job creator?
“How will the funding be balanced? Is this going to be the rich old labs with the rich old guys getting richer, with more postdocs and grad students under them? If so, where will those postdocs and grad students be in 10 years if the initiative is gone? …
“There are too many trainees and not enough tenure track research jobs for them. Many will go on to do other things, things that their scientific training has often ill-prepared them for. And this will continue to happen with the BRAIN initiative, unless the funding is structured to provide for younger investigators.”
We’d like to hear from neuroscientists and other experts on this (we will honor all requests to be anonymous). Exactly how realistic is the project and what should its goals be? Contact us at editors@kurzweilai.net. — Editor
* (Brain Research through Advancing InnovativeNeurotechnologies)
comments 58
by Russell Swanborough
Surely mapping the neurones in a human brain to discover intelligence is about as useful as mapping the transistors in a computer to discover java.
This does not seem like a very useful short term benefit to anyone. It is clearly not going to unlock human intelligence. That is something we already understand and that, I believe, could provide important inputs to this project in order to accelerate the research on areas of particular interest.
by TimothyI
Except that it’s not just positional mapping, it’s functional mapping. The goal is to map the states of various parts of the the brain (at the detail level of the neuron) under various circumstances–sleep, learning, arousal, in disease states and in health, etc. Functional mapping is extremely useful in showing how the brain of an alzheimer’s patient or an epileptic is different from a normal patient. With a large enough sample size it will be possible to determine the precise nature of those diseases, to predict them before they manifest themselves clinically, to determine the comparative effectiveness of various treatments, and to sort out psychological problems from physiological defects.
You’ve set up a straw man (“unlocking human intelligence”, whatever that means). But for more modest goals, a hi-res functional neural map should certainly be the first step. No other approach makes sense–unless your goal is to discourage all federally-funded research. If so that’s a policy debate, not a scientific one.
by Russell Swanborough
@Timothyl
“…You’ve set up a straw man (“unlocking human intelligence”, whatever that means)….”
It is most certainly not a straw man. I believe that understanding how humans interpret afferent information, how they manage and accomplish survival and procreation, and how they generate efferent information to support these goals, are vital to designing human intelligence emulation.
It is what we do and we have done it. It is certainly no mythical concept. We have truly unlocked human intelligence.
by Editor
Russell: could you kindly provide a link to a technical paper, peer-reviewed if possible, that substantiates your claims? What experts have seen and evaluated your finding?
by Sciamtist
Hi Ms Editor
This is a proprietary product with very significant value. Whilst I cannot provide a link to a technical paper, I can answer any questions on capability and, if necessary, demonstrate such capability to any suitably interested persons.
There are two Professors and a number of commercial customers that have been involved in various parts of the research and/or application (but not all of it) and I am sure they would give a positive report on their involvement to date. I would be happy to check to see if they would provide such feedback to you.
Substantiating the claim is as easy as showing it, but I am sure that a few questions that might form the part of what would be needed to demonstrate such capability would be an effective start.
by Editor
OK, feel free to send contact information for unbiased, independent experts in this field and who can provide an objective assessment to editors@kurzweilai.net. Meanwhile, without a peer-reviewed paper or filed patent application or some other form of objective assessment, such unsubstantiated claims do not seem useful for discussion here at this time.
by Sciamtist
Dear Ms Editor
These are not unsubstantiated claims. Please keep in mind that a patent often gives away more than it protects and peer-reviewed papers too easily become public-domain. Our protection is by putting the Human Intelligence Emulation (HIE) functionality into a ‘black box’ to which the user only gets remote access. The software is not for sale.
The reason they are mentioned here is because they do exist and that some very substantial sums of money may be saved by avoiding unnecessary additional research and actually making such capability available.
I am not selling it, I am saying that it is available, just as many other contributors on these threads mention availability of various functionalities without them being labelled as “do not seem useful”. Otherwise we must agree that this is a paradoxical situation: you don’t believe us unless we tell you how it works and we can’t prove it unless we show it to you.
by Editor
I didn’t say I didn’t believe it; I am not qualified to make such a judgment. I said you have not provided substantiated claims. I am not an expert in intelligence emulation, so showing it to me has no value. I suggest you find a qualified independent expert who we both agree is objective, unbiased, and qualified, and who is willing to provide a report to us, and we will be happy to publish this expert’s evaluation. Would you like for me to recommend such an expert?
As for “these are not unsubstantiated claims,” OK, then please state the substantiation here and now, otherwise this is a waste of bandwidth.
by Sciamtist
Dear Ms Editor
To substantiate the availability of these capabilities, we have a number of happy users, including the South African Department of Defence, amongst others.
If you have an expert on HIE that you would like to recommend, we would be delighted to initiate a colloquy.
by Editor
Great. I’ll be in touch.
by Dennis Balson
As mentioned, we need to solve the hard problem of what consciousness is, or if protoconsciousness is a fact, before we learn about secondary consciousness which determines how the brain functions.
by Professor Subbarayan Peri, D.Litt.
Hi,
I am a professor of educational psychology and ‘Learning Scientist’. You know learning science (LSc) is the youngest science. Its locii are: anthropology,computer science, education, linguisitcs, neurology, and psychology. It is closely related to the brain-initiative. Don’t you think so?
Yes! it is.
Learning Science (LSc) is the latest science in the world. The subject matter of the LSc is simple to say, ‘learning’, but really hard to explore. To be more specific, to explore how an organism learns— its (i.e. learning) assumptions, laws, theories, principles, methods, taxonomy, dimensions et. id. The scope of LSc is to understand, measure, predict, and control the process and products of learning (whole of learning) is the LSc. The sciences— anthropology, computer science, education, linguistics, neurology, and psychology are the loci of the LSc. I proposed that both education and research are the two legitimate fields of the ‘science of learning’, their core being ‘learning’. At the same time see the difference; if a content known to somebody is learnt by us, it is designated as is teaching, while a content unknown to anybody in the world up to that point of time that is research. It is surprising, if the two poles of education— teacher and the taught are fused into one pole, it is research! Education and research represent the two hemi-spheres of the brain, the left for teaching (via. reasoning and for reasoning) and the right for research (intuition) respectively. Thus, I want to propose a comprehensive LSc with its two areas viz. i) learning via. teaching (LT), and ii) learning via. research (LR). The two areas are though inter-related, differ with each other in certain characteristics. In this world, to our surprise, only one area viz. LT has been receiving treatment, while LR is yet to. This paper is aimed at to list out the commonalities and differences of the two domains of LSc, and to draw a plan of action for the 21st century. During the first quartile of the century we try to attain the status of a full-fledged science with all its paraphernalia.
It is worth to note that whether theory of research is construed, constructed and established or not there are innumerable research out puts of very high quality from times immemorial on the globe, but the contention of the author is that if the LSc is developed it would be helpful to advancement of the humanity by some centuries by avoiding unnecessary efforts in scientific inventions. Irrespective of my theory of research, Newton established his laws! If my theory of research had developed at that time itself, he would have contributed more than what he did within his life-span! The two hemispheres of brain though found as two parts of the brain, work together in harmony and synchrony. I contend that the functioning of the representatives of them, say learning via. teaching and research should also be work similarly. Of course, they do work as for their nature, but our understanding should be grown to that extent. What is the status of our understanding of LT and LR? is our research question. The answer is simple— we developed the LT and left the RT almost un-attempted.
Of the two wings of the LSc, one wing, the education/teaching has grown to the level of any other advanced systematic science, like physics, physiology, psychology, pathology to be called as a ‘theoretical education/teaching’. While growing itself, it helps to grow the ‘practical education/teaching’. The former stands in the universities as school of education/faculty of education, while its applications live in day-to-day teaching in schools, colleges, and universities. The ‘arts’ phase of teaching is an age old popular belief deep rooted since the Greek history, while the science of education has its existence from Bloom’s taxonomy (1954-57) from the Chicago and Flanders’s class-room inter-action analysis (1966) form the Michigan. The development of the other wing, the research is pessimistic. It has never been treated as a systematic science and as theoretical science. This is the high-time for the educational psychologists to focus on the establishment of ‘systematic science of research’, and ‘theoretical science of research’, as that of education to balance the stability of the mother bird viz. ‘The science of learning’. In my plan, we need to spend at least 10-15 years of this century completing the first quartile. Let’s plan after again, after some time. In fact would have occurred during renaissance or at least 19th century when physical sciences were at spurt of their development or now at least and at last. Yes, now we are late by some centuries!
Dear sir, I have been working on the “Systematic Science of Research”
and ‘taxonomy of research’ since 35 years, developing concepts, tools for objective assessment of quality of researcers and research products. I had discussions, lectures, papers, and books published on the theoretical science of research (just like, theretical physics). In fact, we know only one face of the science of research, the practical or appplied branch it, that too in an informal way.
I request comments, and suggestions on this theme.
Regards,
Professor Peri, D.Litt.
subperi@gamil.com
by Russell Swanborough
Professor Peri
Our organisation has spent over thirty years researching brain learning methods. We have written software that learns just like a human. It will even figure what it doesn’t know and go find it, just like a human. It can research things just because it understands the subject and the concept of use of the subject.
Working with you, it will set your strategy and then ask what you have to achieve it. It will then suggest additions or alternatives where appropriate and, if allowed, manage the actual achievement of your strategy. It will continually learn how to improve this delivery just from knowing what it needs to do and where input information might be found. It will learn how to fix any anomolies and apply them when necessary. Finally, if it is ‘married’ to another similar application that has learned a different environment, then it will immediately blend the two and react accordingly to future challenges.
It cannot unlearn, so-to-speak, thus it will quickly out-remember any human. The singularity could be close.
by Professor Subbarayan Peri, D.Litt.
Can I have more details and contact of your org?
by Sciamtist
You can contact us at info@sciam.co.za.
by Publius
The visible brain has arrived: Scientists at Stanford University reported on Wednesday that they have made a whole mouse brain, and part of a human brain, transparent, so that networks of neurons that receive and send information can be highlighted in stunning color and viewed in all their three-dimensional complexity without slicing up the organ. Even more important, experts say, is that unlike earlier methods for making the tissue of brains and other organs transparent, the new process, called Clarity by its inventors, preserves the biochemistry of the brain so well that researchers can test it over and over again with chemicals that highlight specific structures within a brain and provide clues to its past activity. The researchers say this process may help uncover the physical underpinnings of devastating mental disorders like schizophrenia, autism, post-traumatic stress disorder and others.
The work, reported on Wednesday in the journal Nature, is not part of the Obama administration’s recently announced initiative to probe the secrets of the brain, although the senior author on the paper, Dr. Karl Deisseroth at Stanford, was one of those involved in creating the initiative and is involved in planning its future.
http://www.nytimes.com/2013/04/11/science/brains-as-clear-as-jell-o-for-scientists-to-explore.html?hp
by Sean Stoneburner
For those who are interested, I created a Google Plus Community Page for the BRAIN Initiative. https://plus.google.com/u/0/communities/102531740760787173202
by Medusa
Somebody tell them that we need to understand the physics of emotion/consciousness. There is a huge hole in our understanding of the physics of the universe in this subject.
by Editor
Could you expand on that?
by H1dra
I’d also like an elaboration. However I’d suggest you first check out resources of the likes of neurohackers dot com and second develop a sensitivity to the social and individual patterns of your emotional spectra. It is a lifetime long study…
by Pete
Well, emotions are well-explained – they are a primordial type of intelligence (for preparing an individual into emergency situations, switch between a sympathetic-dominant state and a parasympathetic-dominant state, etc).
by H1dra
I don’t get it Editor… praising Pete is somehow a personal attack on everyone else or something?
by Pete
How is my comment a “personal attack”, or even in any ways related to that?
Or, do you mean to praise me? Then well, thank you.
by H1dra
Yes. My post was not approved.(it’s the new ‘delete’)
Consider yourself meta-praised, lol :-)
by Editor
I don’t know what you are referring to.
by Micahel B
As a student working in the cognitive sciences and AI, I would first like to point out that the general goal presented, namely the mapping of the brain’s connections (primarily neural, but also functional), is a necessary step forward for the science of the mind (in addition to related sciences, i.e. computer science).
There are more specific research goals already emerging, both short term and long. Short term, I’ve seen proposals to refine calcium-based imaging of neural firing. Long term, such methods are insufficient for a number of reasons: what is required are methods (singular, or in composite) that can capture system wide firing throughout the brain at the spatial resolution of the neuron, and the temporal resolution of a single firing. This would be the holy grail of the project. A more approachable goal is to map the connections between neurons. This project is NOT invalidated by the fact that all brains have different (and constantly changing) connections, because what is of interest is the general probabilistic structure of the brain (the brain is inherently probabilistic), deviations thereof, and their correlations with behavior.
Whether it is achievable in the next 10 years is up in the air (I would bet that the structural map is), but I would say that such technology is easily worth $100M.
by trakk
Every project has its skeptics.
by TimothyJ
Exactly–this is mindless pessimism. Yes, there are trillions of connections, but we have the technology to handle that kind of data today, and our capability to handle big data is increasing exponentially.
Also, the PATTERNS of connections are what’s important. If we develop the ability to map connections in a normal person, and compare them with a person who has a brain abnormality, the basic, core features of that abnormality will be revealed, without all the noise that comes from trying to diagnose via clinical behaviors and symptoms. It may allow early (pre-clinical) diagnosis and treatment, something that almost always improves outcomes.
This is similar in concept to genome mapping–everyone has a unique genome, with hundreds of millions of base pairs, but by using big data to compare thousands or millions of genomes, associations with physical and medical abnormalities can be discerned. This is the whole idea behind GWAS (genome-wide association studies)–a relatively new technique that is a major focus at the NIH, because it will produce huge benefits in early diagnosis and treatment. It would be impossible without the basic work performed in the Human Genome Project–something that was seen as a boondoggle and “full of empty promises”.
by Pete
Given that technology in many fields are growing exponentially, and that we may potentially use quantum computers in this (and other similar) project, I am very optimistic.
The problem is that, too little money is spent in this project.
by Publius
Meanwhile, the Europeans are spending $1 BILLION on their “Human Brain Project,” which will “integrat[e] everything we know about the brain into computer models and us[e] these models to simulate the actual working of the brain. Ultimately, it will attempt to simulate the complete human brain.” HBP will incorporate the work of Henry Markram’s Blue Brain Project, among many others. One again, as with the LHC versus the Superconducting Super Collider, the USA is left in the dust on the most important fundamental reaearch by short-sighted politicians. Is the EU really in a better economic position than the US? Hardly. But they see the future and want to bring it on, while the US looks at its feet.
by Joel C.
I would blame it on Americans’ myopic interest in Kim Kardashian. I know I’m overgeneralizing but I’m speaking for the majority.
by Pete
It seems that more people in EU are Singularity-minded. I hope that is accurate description.
by Russell Swanborough
And that’s the problem. It will *simulate* the human brain. We are already emulating the human brain and can provide forward advice on what to look for.
by d
my greatest fear i government involvement. They literally screw up anything they touch. Private industry will find a reason for this and there will be profit. This is why we have Iphones and Ipads. Where as, with the government we have deficits equal to 50,000 for every man woman and child even the space program that took us to the moon is in shambles now. Its like the revers Midas touch they will tie up money and good scientists producing a pile of turds. Even with all the money wasted on the human genome project the government had to concede that Craig ventner beat them to the punch for no tax dollars spent.
by melajara
iPhone and iPad is hardly a metric of any significant achievement, very poor choice IMHO. Those are just nice (but not outstanding) design examples leveraging on a 50 years old paradigm : downsizing.
But in Apple case, the design achievement is overcompensated by the encompassing locked in syndrome (OS + market control) to milk the (consumer) cow.
The irony of this greedy scheme, Apple is sitting on a stupendous $170 Billion cash pile, seemingly without any c(l)ue how to intelligently spend it.
Now contrast this with what I.B.M is making for the advancement of science and technology with 1/3 of Apple resources, very telling too.
by Cybernettr
Actually, I read in Macworld it’s “only” $120 billion.
Maybe we’ll see some phenomenal advancements in Apple’s Siri technology in the near future. Who knows?
by Pete
This is a good counter-example of some Singularitarians’ belief “all technologies lead to Singularity”.
Not all technologies lead to Singularity.
We need a clear goal.
That goal, in my opinion, is creating a Whole Brain Emulation, and from that, create human-equivalent artificial intelligence.
If we learn to create human-equivalent AI, it will be easy to create super-human AI.
by Editor
Yes, that’s a good partial summary of some of Ray Kurzweil’s concepts, but the neuromorphic premise is just one of several approaches to achieving the Singularity (assuming such a goal is possible), and there’s no definitive evidence yet to exclusively support that particular premise, which has been criticized by some experts. Also, there’s no consensus on how detailed (neuron level? neuron + glia level, as Stein suggests? protein level? electron-level? quantum level? etc.) such a brain model would need to be in order to be useful for development of intelligent machine systems.
Other Vingean Singularity scenarios that may deserve study include human intelligence augmented by computer interfaces or bio/nano enhancements, and large computer networks (and their associated users) that “wake up” as a superhumanly intelligent entity. (http://www-rohan.sdsu.edu/faculty/vinge/misc/singularity.html)
As for the goal of the BRAIN initiative, as stated by its funders, it
has nothing to do withis not directly related to the Singularity. Its focus is on a “research effort to revolutionize our understanding of the human mind and uncover new ways to treat, prevent, and cure brain disorders like Alzheimer’s, schizophrenia, autism, epilepsy, and traumatic brain injury,” according to a statement by the White House, http://www.whitehouse.gov/blog/2013/04/02/brain-initiative-challenges-researchers-unlock-mysteries-human-mind, which is why President Obama was introduced by the NIH director.In contrast, the EU’s Human Brain Project includes “a completely new category of low-energy computing systems, with brain-like intelligence” as a goal, mentioned after the two goals: “gain fundamental insights into what it means to be human and develop new treatments for brain diseases….” (http://www.humanbrainproject.eu/files/HBP_executive_summary.pdf).
If you’re looking for a project that’s based exclusively on the neuromorphic model, the goal of DARPA’s Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) program, currently being implemented primarily by IBM, is to develop electronic neuromorphic machine technology that scales to biological levels.”
(http://www.darpa.mil/Our_Work/DSO/Programs/Systems_of_Neuromorphic_Adaptive_Plastic_Scalable_Electronics_(SYNAPSE).aspx)
by Pete
Thank you, Editor. The information and links you provide are helpful.
by Editor
Good. I’ll be interested to learn what new approaches you find. Keep us advised?
by Russell Swanborough
Pete, we have Whole Brain Emulation already announced on 11 April. Please call with questions.
by anon
All the other posters here make a good point, but it is still highly important that the GOAL of the project is a good one. One of the larger reasons efforts like the moon landing and the human genome project were so successful was that the stated goal inspired people and made them interested, made them care.
Things like this need a trajectory, a start and and end point. We won’t get any of the (no doubt fantastic) unexpected payoffs with an unclear or impossible to measure end point.
I’m not saying a good goal can’t be found, but a little less brainless singularian fanboying and a little more appreciation of the role of valid structured criticism in the scientific process won’t hurt.
by H1dra
Perfect post, anon. I’m afraid I’ll do only disfavor by associating my name with it.
by Rob
Boondoggle. As much of a waste of money as the previous administration’s attempt to get people interested in space again. Rather than deal with real issues, they spotlight pie-in-the-sky initiatives like this. Scicurious has a point and illustrates the weakness of any government-sponsored endeavor. How do you even know there will be a use for this?
For those of you who think that throwing money at science will yield unanticipated benefits, you’re wrong. There is nothing to suggest that anything good will come out of this. There are much more promising endeavors out there which could use this funding that can describe what benefits will accrue if they get funded. Unlike this, which is why it’s a boondoggle.
by H1dra
Exactly. But careful =) they don’t like so called “naysayers” around here…
by Andy
Obama and the government know how to spend your money better than you do. Except, of course, when it’s something you disagree with.
Heaven forbid that Obama and the government allow people to spend and invest their money as they chose. Rather than everyone using the government to force everyone else to pay for what I want or you want or your neighbor wants, even though we may all disagree; everyone uses the force of government as their enforcer for their pet projects.
by H1dra
I mostly agree with your point. I hope kickstarter will soon expand to scientific enterprises.. sens will rocket forward then! And maybe we can save our older loved ones.
by Shaun M.C.
It’s a bit premature to call this a boondoggle given the fact that none of the agencies issuing grants have announced how they intend to issue the allocated money.
by H1dra
Interesting.
is there a way to prevent waste depending on the actual policy of such an allocation?
I’m afraid once they say they _will_ fund it, anything except perhaps for the uttermost failure, would not be enough to deter further allocation of already set in stone funding at the outset…
if this could indeed be done honestly, which I doubt, then I’m all for it.. though I still prefer sens =)
by Tom B.
As with any stretch scientific objective, the small problems along the way often yield huge unanticipated dividends. Maybe this project will yield better brain scanners that can do things we can’t even guess. Maybe it will produce software that leads us to be able to reverse engineer a brain. Maybe it will cause the create of computer systems that can open the doors to low cost self driving cars or…who knows?
The point is: “the road is better than the inn.”
by bernard garner
Somehow this is all reminiscent of the pessimistic hand wringing that went on at the begining of the human genome project. I also suspect a lot of the whiners are looking enviously at the project budget and thinking just how much more deserving their own projects are.
by H1dra
Every high budget project is bound to succeed then? Question is, how efficient is to have vague goals and there are more efficient ways to go about it? Maybe those more efficient ways already exist, yes, but not necessarily.
by Pete
I believe that the clear goal should be “to create a Whole Brain Emulation”.
WBE leads directly to creating a human-level artificial intelligence, and the Singularity.
by Pete
What is even more frustrating is that the world’s governments and other groups spend billions (and sometimes even trillions) on unnecessary tasks (wars, entertainments, and many others). If those money are spent on science (especially AI), the world would be a much better place.
by Alex
Moaning and handwringing that will be invalidated as the project makes great strides these next few years.
by CLains
exactly my thoughts. as if they had any clue on how to land a man on the moon in 1960
by Pete
I agree with you. Everyone must remember Clarke’s First Law.
by H1dra
Alex, it is hard not to get ANY thing out of billions of dollars. One would have to try real hard.
However, we don’t just want to get ANY thing, we want to get as much as possible. Don’t you agree?