The Artificial Intelligence Problem (or not)

Yes, I understand that DARPA isn’t the only one doing R&D. As a percentage of the budget though, we are spending less every year on R&D. That’s not missing creep, it’s choosing to spend money on other things.

US Federal Research Spending Is at a 60-year Low. Should We Be Concerned?

Sorry, I might not have been clear. My point is that putting money in DARPA for R&D isn't needed anymore because I think the ROI is probably bad. Democratizing the R&D capability down to lower levels is more effective, and then creating expediated contracting vehicles like what you can do with DIU is the better way to go. Just let industry do it for us and limit DARPA and IARPA to color in the corner for things that maybe my grandchildren can benefit from.
 
Sorry, I might not have been clear. My point is that putting money in DARPA for R&D isn't needed anymore because I think the ROI is probably bad. Democratizing the R&D capability down to lower levels is more effective, and then creating expediated contracting vehicles like what you can do with DIU is the better way to go. Just let industry do it for us and limit DARPA and IARPA to color in the corner for things that maybe my grandchildren can benefit from.

I’m not a financial person or a comptroller, but we (the DOD) pay for things via different colors of money (MILCON, Operations and Maintenance, R&D, Procurement, etc). So even asking industry to do it still requires funding, which in this case would be R&D funds; which we are spending less and less on. Industry isn’t going to do it for free. Sure they might do this in some cases, but industry isn’t in the business of doing stuff for free, otherwise they won’t be around very long.
 
I love DARPA. I wish we could have a bigger R&D budget. Even if shit fails there’s always a takeaway that can be applied elsewhere. Our D&D budget has shrunk by like 80% compared to what we spent in the 1950s.

Sorry, I might not have been clear. My point is that putting money in DARPA for R&D isn't needed anymore because I think the ROI is probably bad. Democratizing the R&D capability down to lower levels is more effective, and then creating expediated contracting vehicles like what you can do with DIU is the better way to go. Just let industry do it for us and limit DARPA and IARPA to color in the corner for things that maybe my grandchildren can benefit from.

That is what DARPA already does. They fund a ton of research in universities. My wife’s doctoral thesis was funded by DARPA, and she made implantable 3-D printed skin scaffolds for burn patients. Industry has no desire to fund an obscure thing like that with a low ROI. The government does because guys were getting burned like crazy for a decade in Iraq and Afghanistan.
 
Back
Top