One of the buzzwords popping up all over the place now is >grid computing=, where a group of computers are wired together for what is usually either a scientific computing task or a graphical rendering of seriously pissed-off dinosaurs, turbine powered air racers, or exploding death stars.  Oracle Corp. (client/server databases) has a grid version of their database environment, some computer whizzes at University of Illinois have assembled a cheap supercomputer out of 50 Sony Playstations, and projects at the University of California have wireless sensors arranged in a grid so that data acquired at any one of them can >hop= from one wireless node to another until it=s received by a data collection and visualization station.

A database of all known insect species would be substantial, perhaps to the extent that a grid-style server farm would be essential.  Insect traps would be placed in harbors and airports and in areas known to be ecologically rich.  Each sampling point would periodically scan whatever flew in, and then attempt to locate whatever it was in the database.  A collection of samples taken at several stations in a particular region might consistently fail to return an existing species, although each query might have quite a bit in common with its >neighbors=.  The presence of such a collection of records would hint at a previously unidentified species.  Naturalists would presumably converge on the site to see if they can figure out what=s going on.

Another feature of the 1950's was a paranoid fear of the >enemy=, an emotional state some of us are in now.  In the Cold War era, our reaction to this was to pour money in to jet engines, aircraft R&D, missiles, nuclear weapons, and a bit later on, spy satellites.  Our reaction now is to watch everything, from our airports and harbors to the rugged terrain between Pakistan and Afghanistan to tramp steamers that seem to wander around various ports in Southeast Asia and Australia.  This watching of course goes on through a network of imaging satellites, telecommunications intercepts (Echelon) and piloted and pilotless spy planes.

 Like much of the spending in the 1950's, today=s defense applications are trial and error and have varying degrees of success ranging in some cases down to none at all.  It may take ten or twenty years to teach computers to reliably find one needle in a farm full of haystacks, and by that time the technology may be primarily civilian.  Many of the miliary users of what we have today will be in the civilian workforce by that time, probably making a nice living with what they know.

One of the things we=re looking for is a release of infectious agents.  Terrorists might be able to kill a few city blocks worth of people with an anthrax outbreak, but the forces of nature reduce this to triviality.  Nature has billions more chances to bring a nasty into the human community than anyone with a lab full of petri dishes.

Grid sensors (i.e. a collection of thousands of data collection elements on one chip) connected to grid messaging elements (hopping the data from one collection point to another) to a grid database and then to a grid event filtering and recognition system is an obvious technical solution.  Such a system would pay for itself were it able to limit one epidemic or crop infestation.  We assume that with the proliferation of global commerce that such infections are more likely to appear, and once they do, spread with lightning speed.

The SETI (Search for ExtraTerrestrial Intelligence) project uses grid computing, in this case the grid being people who volunteer the use of their PCs over the internet.  Some of the free ISPs require that their users leave their idle computers connected so that the ISP can run grid computing jobs.  The latter is certainly a fair proposition as long as the job being run wouldn=t cause the subscriber any offense.  >Flat earthers= might have a problem with SETI, but they might have even more problem with geophysical tomography.

What does one do with 1,000,000,000 teraflops of computing power?  Supposedly unbreakable ciphers do give way if enough computing power is thrown at them.  This is how we got into Al-Qaeda=s word processing documents in Kabul.  If a code takes 500 years to break with a super computer, how long does it take if 100 million PCs are invited to help out?

One of the big challenges now is finding the one compound out of billions that might make a safe and effective drug.  Let=s say we find a bug... so what?  What we need is the remedy.  Drug research is, in many respects, molecular modeling, which looks at chemical interactions.  Chemical interactions are often heavily influenced by >folding=, the way that carbon chain molecules wrap themselves into helices and other compact shapes.  Billions of compounds applied to billions of potential interaction sites on human cells: serious work.

I=ve never seen a mosquito trap that works particularly well, so it=s time for an array of sensors to be placed in my back yard along with some computer controlled laser guns.  If the cameras see a mosquito they zap it, first making sure I=m not standing behind the intended target.  This might take more oomph than my desktop PC, but if yours isn=t doing anything anyway...