Wednesday, May 27, 2020

The National Science and Technology Foundation?

A proposal is being put in front of Congress that would reshape the National Science Foundation into the National Science and Technology Foundation.  The Senate bill is here, and the House equivalent bill is here.  The actual text of the Senate bill is here in pdf form.   In a nutshell, this "Endless Frontiers" bill (so named to echo the Vannevar Bush report that spurred the creation of the NSF in the first place) would do several things, including:
  • Create a Technology Directorate with its own advisory board (distinct from the National Science Board)
  • Would identify ten key technology areas (enumerated in the bill, initially (i) artificial intelligence and machine learning; (ii) high performance computing, semiconductors, and advanced computer hardware; (iii) quantum computing and information systems; (iv) robotics, automation, and advanced manufacturing; (v) natural or anthropogenic disaster prevention; (vi) advanced communications technology; (vii) biotechnology, genomics, and synthetic biology; (viii) cybersecurity, data storage, and data management technologies; (ix) advanced energy; and (x) materials science, engineering, and exploration relevant to the other key technology focus areas)
  • Would have funds allocated by program managers who may use peer review in an advisory role (so, more like DOD than traditional NSF)
  • Invest $100B over 5 years, with the idea that the rest of NSF would also go up, but this new directorate would get the large bulk of the funding
This article at Science does a good job outlining all of this.  The argument is, basically, that the US is lagging in key areas and is not doing a good job translating basic science into technologies that ensure international primacy (with China being the chief perceived rival, though this is unstated in the bills of course).  If this came to pass, and it's a big "if", this could fundamentally alter the character and mission of the NSF.  Seeing bipartisan congressional enthusiasm for boosting funding to the NSF is encouraging, but I think there are real hazards in pushing funding even farther toward applications, particularly in a governance and funding-decision model that would look so different than traditional NSF.  

It's worth noting that people have been having these arguments for a long time.  Here is a 1980 (!) article from Science back when a "National Technology Foundation" proposal was pending before Congress, for exactly the same perceived reasons (poor translation of basic science into technology and business competitiveness, though the Soviets were presumably the rivals with whom people were concerned about competing).  The NSF has their own history that mentions this, and how this tension led to the creation of the modern Engineering Directorate within NSF.  

Interesting times.  Odds are this won't pass, but it's a sign of bipartisan concern about the US falling behind its technological rivals.

7 comments:

Anonymous said...

If the NSF takes a turn towards applications, basic sciences are going to really suffer. This is really a long term vs short term discussion.

Anonymous said...

Advanced energy seems like something out of a certain publisher's playbook.

JDeibel said...

Every time I read something about this proposal, I oscillate back and forth multiple times as to whether or not I could get on board with this. There has to be some basic protections if not deliberate firewalls built into this to protect not only supporting basic science but also NSF's peer review driven process. While there is always room for improvement with NSF's review process, I haven't found any instances of something better than it. When I first heard about this bill, my instant thought was that it would be great to have pathways to apply science and develop technology that was not the SBIR/STTR pathway and these new pathways could support scientists and engineers who has proven ideas and potential applications. To this day, I am still amazed at the lack of robust and consistent review in the SBIR/STTR programs, with the exception of the NSF programs. So, then I go and read more about this bill tonight,and I learn that the tech programs would not have to institute peer review.

For this to be viable in my opinion, the new tech directorate would have to have the same peer review policies that exist now at NSF and the bill would need to make certain that basic science funding would remain intact and not have the potential of getting raided by the tech part of the new agency.

Like you said Doug, I don't see this getting passed anytime soon. It will have to pass in the Senate and I just don't see the majority supporting growing a government agency by this much. I think that they would be inclined to take the funding $ from this and directly send it to tech companies. Would that help advance technology development in our country, most likely no, but since when did that kind of thing hamper the senate?

Anonymous said...

The peer review grant process is precisely why scientific progress and subsequent translation into technology is so slow with the NSF. Most scientists, including myself, are poor judges of feasibility and impact of the truly innovative stuff NSF should be funding. Ironically, I think China is often even worse in this sense since they are very focused on misleading citation metrics. On the other hand, they also like to bankroll high profile scientists to achieve their scientific vision.

That said, the alternative is to have visionary leadership who can weather criticism and nurture truly groundbreaking innovation. Unfortunately, those types of people are very rare, particularly so in basic science administration where no one wants to work in the first place.

Anonymous said...

Peer review may have problems (slow, risk averse, some randomness because of opinionated reviewers without real expertise), but the DARPA/DOD approach presumes that all program officers are visionary technical experts with excellent judgment. When you put complete authority on program officers, especially rotators, you basically encourage the old-boy network. People can have their favorites, and large sums of money can get allocated based on personal relationships rather than critical technical evaluation of proposals.

Grumpy said...

"Peer review may have problems (slow, risk averse, some randomness because of opinionated reviewers without real expertise), but the DARPA/DOD approach presumes that all program officers are visionary technical experts with excellent judgment. When you put complete authority on program officers, especially rotators, you basically encourage the old-boy network. People can have their favorites, and large sums of money can get allocated based on personal relationships rather than critical technical evaluation of proposals."

That's really an excellent comment. I see grant peer review the same way.

Anonymous said...

You need both, I'd say at least 30% should go into the DoD way of doing things.