I Want To Be a Mad Scientist

I Want To Be a Mad Scientist

I Want To Be a Mad Scientist

I Want To Be a Mad Scientist

I Want To Be a Mad Scientist
Innovation, the Internet, gadgets, and more.
Aug. 9 2007 3:36 PM

I Want To Be a Mad Scientist


Illustration by Robert Neubecker. Click image to expand.

ANAHEIM, Calif.—The exhibit hall here at DARPATech may have its serpentine robots, morphing aircraft, and ultrasonic blood clotters. But for the really weird stuff, you have to go to the lectures.

Daniel Engber Daniel Engber

Daniel Engber is a columnist for Slate

In the grand ballroom, amid flashing blue lights and rock guitar riffs, 50 or so program managers take turns laying out their goals for the next two years of DARPA funding. The guy in charge of materials science wants to support research into something called "programmable matter." Eventually, he says, we'll have a blob of goo that can form itself into a hammer in one instant, a wrench in the next. "It's an instant toolkit … a universal spare part!" Another manager shows trippy slides of veinous bio-scapes, and wonders how we might "killproof" our troops in Iraq and Afghanistan. He's looking for proposals to create a universal immune cell that can be used to protect soldiers from any deadly pathogen. One by one, the DARPA administrators sketch out their nutty techno-fantasies for an audience of 3,000 professors and engineers. And they've got billions of dollars to make those fantasies come true.

Darpa. Click image to expand.
DARPA director Tony Tether speaks to the crowd in the grand ballroom

Let me repeat a question I asked yesterday: How do you get this job?

I head over to the employment desk in the exhibit hall to find out. Good news—DARPA's hiring. A human resources rep hands me a branded piece of chocolate and suggests that I mail my résumé, cover letter, and radical, world-changing idea to employment@darpa.mil. (I can even use this PDF template to lay out my futuristic proposal.) This feels a bit like applying for double-O status on Monster.com. If I can land this advanced military research gig, they tell me I'll be making up to around $160,000 a year, with five weeks of vacation and a $100 monthly allowance for the Metro.

Back in the ballroom, I watch the program managers exchange high-fives and fist bumps as they trade microphones between speeches. These are mostly academics or former industry scientists in their 30s or 40s who have taken a leave from their regular jobs to work for DARPA. Each gets just four or six years to push their high-risk research programs as far as they can go. The agency considers this rapid turnover—25 percent of the staff per year—the main ingredient in its "secret sauce" for technical innovation. When a new manager comes in, she can build off of earlier research programs, ignore them, or even work to undo them. (In that sense, DARPA functions a bit like the White House.)

It's a funny way of doing business when you're focused on highly speculative, long-term innovation. How far you can get toward making programmable matter or universal immune cells in four years? Is it worth investing DARPA's $3 billion annual budget in these short bursts of pie-in-the-sky research, or could our research dollars be spent more efficiently elsewhere?


The Defense Department uses DARPA to spread out its investments, just like you would with a 401(k) plan. Most of the research money flows into reliable projects that produce slow and steady returns. But about 15 percent to 20 percent goes toward the high-risk, high-yield projects on display here in Anaheim. Sure, a bunch of cash might get flushed away on ill-advised programs, like the mechanical elephant and the terrorism futures market. But every once in a while, you hit the jackpot.

It's hard to know if the Pentagon puts too much money into these wild projects—perhaps 5 percent or 10 percent of the budget would be a more reasonable outlay. But it does seem like every department or government agency could benefit from having its own nutty research arm. The Department of Energy would be an obvious choice; the National Academies have already suggested that ARPA-E could help rescue us from the coming environmental catastrophe. Meanwhile, DARPA funds research to protect soldiers from tropical diseases and traumatic injuries but ignores far more important problems like cancer and aging. How about an ARPA for the National Institutes of Health? A NOAA-ARPA could help to prevent the next Hurricane Katrina, and an advanced research division of the Library of Congress might bring the Google book project to fruition. And you can just imagine what would happen if we gave an ARPA to the U.S. Postal Service. (Hello, video stamps!)

For now, the sci-fi approach to government spending seems limited to the creepiest branches of government. Starting next year, the nation's intelligence agencies will open their own advanced research franchise. God knows what innovations this will yield in the field of torture technology. Maybe someday we'll have a waterboarding robot or a new way to "leak-proof" our government surveillance programs.



Acronym Update: Here are a few more linguistic causalities in the Pentagon's GWOT, culled from my ongoing Acronymic Stupidity Survey of the conference exhibit hall:

• Countermeasures for High Altitude Nuclear Detonation (Sleight of HAND). OK, "HAND" works, but what's the deal with "Sleight of"? And what does this project title even mean? Is DARPA just slipping these nuclear detonations under the table?

• Steep-subthreshold-slope Transistors for Electronics with Extremely-low Power (STEEP). First of all, you can't create an acronym that's identical to one of the words in the project name. (I have a similar gripe with the Threat Agent Cloud Tactical Intercept & Countermeasure, or TACTIC.) Second, I shouldn't have to comment on the missing letters from subthreshold, slope, for, with, and low. Let's call this project like we see it: SSSTFEWELP.

• Semiconductor-Tuned High-Temperature Superconducting Filters for Ultra-Sensitive RF Receivers (SURF). No acronym matches this one for shameless word-pruning. It's almost impossible to tell which letters in SURF go with which words in the full project title. Is it "Semiconductor … Ultra … RF?" "… Superconducting … Ultra … RF"? Or maybe it's just "… SUpeRconducting Filters …"

And the big winner:

• Predictive Analysis for Naval Deployment Activities (PANDA). The words in the title seem unforced—you can almost imagine that the cuddly animal acronym was just a happy coincidence. And plenty of bonus points for the project's psychedelic logo, a panda surfing the barrel of a sick wave.