There’s a great article on the future of super-computing in Wired (click on the link above).

Money quote:

“Amid the beckoning fantasies of futurism, the purpose of whatever comes next – like that of today’s petapede – will be to serve the ultimate, and still the only general-purpose, petascale computer: the human brain. The brain demonstrates the superiority of the edge over the core: It’s not agglomerated in a few air-conditioned nodes, but dispersed far and wide and interconnected via myriad sensory and media channels. The test of the new global ganglia of computers and cables, worldwide webs of glass and light and air, is how readily they take advantage of unexpected contributions from free human minds, in all their creativity and diversity. Search and you shall find.”

This year, for the first time in a while, I’ll not be at the Society for Neuroscience meeting (which starts tomorrow in Atlanta). This is a very large scientific meeting (typically roughly 25,000 folks register) and for my field of neuroscience, it’s really the only place where everybody in brain sciences gets together in one venue.

And it’s not just the scientific sessions–I’ve always also enjoyed tremendously the opportunity to get together with colleagues and friends who are at other institutions, but with whom ties are still very strong.

So why I am staying home at Krasnow?

Fundamentally, the reason is that with the addition of our expansion space, multiple on-going faculty searches, and the nearness of the operational starting-line for our brain imaging center, there’s just too many Krasnow-specific operational details that require close attention.

In the meantime I wish all of my colleagues an excellent meeting and I’m very much looking forward to next year in San Diego.

Going beyond deception detection, neuroscience may soon be able to play a more practical role in counterterrorism. The enabling technologies are non-invasive brain imaging (particularly with data-fusion from multiple imaging modalities), trans-cranial magnetic stimulation, virtual reality environments, and the terrabytes of newly shared neuroscience data indexed in massive databases.

The goals might be: 1) to gain insight into the brain states that are necessary and sufficient to subserve a terrorist act (particularly when the act results in the suicide of the perpetrator), 2) to use technologies to modify such brain states in a controlled fashion–in the laboratory first and subsequently the field and 3) to explore the use of artificial sensory inputs to transition brains from terrorist-permissive to the normative condition.

Ultimately, terrorism is a human act produced as a result of neuronal actions in a human brain. While human acts are often very complex in nature, many behaviors have been explained quite well at the neuronal level (e.g. addiction and drug abuse). If the neural basis can be understood, then interventions can be devised.

I had the pleasure yesterday of visiting HHMI’s new Janelia Farms Research Campus for an open house. Located just north of Dulles Airport’s runways along the Potomac River, the facility is truely a marvel. I had visited the site during the initial construction phase–the finished facility is really both an architectural landmark and also potentially a new international nexus for neurobiology.

On the virtual side, I had my first tour of the new Allen Institute’s mouse gene atlas which in its own way is equally extraordinary.

Money quote:
“By mid 2006, the Atlas project will have completed mapping gene expression in approximately 20,000 genes all of which will be available through our site.”

Go check it out, they’re all there.

Jim

There is an important editorial in this week’s Journal of Neuroscience about the changes that are affecting how science gets disseminated. While I don’t necessarily agree with the authors, I do feel that change is in the air–ten years from now, I think scientific journals will look very different from the way they do today.

The key point that is missed by many of the folks who are advocates of open access is that there is a value-added component to a publication that arises out of the editorial process (all the way from peer review to archiving) that costs real money. If the money isn’t going to come from subscriptions (open access) then it must come from somewhere else. Most current notions of a solution come down to page charges of some sort.

The question is can those real costs be lowered using technology? Certainly at my own journal, we are finding that our new on-line editorial system is making the process much easier (and hence probably less expensive). But there are aspects of the editorial process which are don’t seem to have much technological cost elasticity (e.g. peer review). I don’t see how a robot reviewer is going to replace a human scientist any time soon.

Jim

Congratulations to our nearby colleagues at the Carnegie Institution of Washington for the news this morning that Andrew Fire is sharing 1/2 of the Prize for work he did while at Carnegie on RNA interference.

Jim

Today the science leadership of Krasnow took lunch together (as we do every semester) at a wonderful Turkish restaurant down the road. The conversations were substantive ranging from Rob Axtell talking about using agent-based modeling in the context of theoretical medicine to Avrama (Kim) Blackwell expressing the hope that we would one day, really understand the cellular basis of both classical and operant conditioning. There are now 17 PI’s at Krasnow–there were perhaps five, when I took the job of director in 1998. With a total staff of around 60, the Institute has really grown.

Jim

One of our own will be speaking soon at Mason’s Vision Series–a series of lectures open to the public. If you live in the National Capital Area, consider putting it on your calendar.

Towards Virtual Brains

Giorgio Ascoli, Krasnow Institute
Monday, October 23, 2006 at 8 pm
Center for the Arts Concert Hall

For decades, the construction of a computational model of the brain has been a kind of “holy grail” in both brain science and computing. This lecture offers an overview of the architectural principles underlying the complexity of the human and mammalian nervous system, and how they relate to electrical activity and cognitive function. Following this introduction, we will consider how computer simulations based on detailed experimental data can advance our understanding of how the brain works, with potential benefits for both biomedical and computational science.

Jim

Tomorrow evening promises to be an interesting one, even for a place as unique as the Krasnow Institute. Faculty from the George Mason philosophy department and others with like-interests will enjoy a wine and cheese gathering with their Krasnow colleagues to discuss different approaches to understanding and studying “mind”. We’ll gather tomorrow evening in the Krasnow Great Room at the first of several year long events to celebrate the Institute’s scientific program. Harold Morowitz, Clarence J. Robinson professor of Biology and Natural Philosophy will get the event started with some preliminary remarks and then the rest of the time will be spent talking about mind–often across great disciplinary divides. Should be fascinating.

Jim

One of the challenges that we face at Krasnow is matrix management. Essentially Krasnow is a project-driven sort of place, but our faculty also report to function-driven line managers (also known as deans and department chairs). There are lots of models for matrix management–but they all seem to require a bit of tolerance for ambiguity. Seems to me that this is one of the prices we pay for being an integral research unit of a large university.

On the other hand, our sister stand-alone institutes, have to worry about funding their entire program from endowment and soft money. That’s a recipe for a bit of ambiguity also.

The real key to successful matrix management is a close collaborative relationship between the project managers and function managers on issues related to evaluations and raises.

Jim