I'm showing my internet age in thinking that the right image to put at the top of this post is either the Drudge Report emergency light icon or an animated Star Trek "red alert" sign.
This is a funding trajectory that has not been seen since the 1970s. Now, because of budget uncertainties and disruptions last year, there was a big burst of activity late in FY25, and eventually the NSF did end up spending about what it was budgeted. I spoke with one program officer at NSF last month who said that they fully intended to get there again this year, even if it meant he didn't have a vacation until September.
As you may be aware, NSF spending is incredibly low this year. How low is "low"? Check out this graph from grant-witness.
This is a funding trajectory that has not been seen since the 1970s. Now, because of budget uncertainties and disruptions last year, there was a big burst of activity late in FY25, and eventually the NSF did end up spending about what it was budgeted. I spoke with one program officer at NSF last month who said that they fully intended to get there again this year, even if it meant he didn't have a vacation until September.
A lot of people had looked at the trajectory above and worried that we are headed toward some kind of very bad outcome. For example, if NSF is underspent by $3B by August, whether because of direct OMB opposition or because the award office at NSF is told by political leadership not to make awards, then it might be nearly impossible for NSF to spend its budget, at which point there could be a pocket rescission. Basically, the executive branch has wanted to enact 50+% cuts to NSF; Congressional appropriators have said "no", but the executive branch may be trying to get their cuts anyway. This would a terrible precedent. If it happened you might expect Congress to be upset that their appropriations were being ignored. There would likely be lawsuits.
Today, however, this story broke in Science. Supposedly, there are going to be broad cuts to many parts of the NSF, at the level of 20-30% in the present fiscal year, despite the fact that the NSF budget is only down 3% from last year and there is statutory language in the appropriation bill saying that no directorate could be cut by more than 5%.
The article basically says that it is likely that the funds are going to support the X-Labs effort run out of the TIP directorate. What is an X-Lab? I have some inkling because I attended the webinar about the present solicitation a couple of weeks ago.
The idea of X-Labs comes from proposals like this. The basic premise is (1) The present system holds back innovation for some and we need to be more flexible and entrepreneurial. (2) We could bring together teams of people who could be in a position to do something transformative, with definite technology applications, but whose work is at an early stage such that it's too low a technology readiness level to attract VC/angel investors who could support a startup, or is too far off from deployment to be partnered with industry as in the long running SBIR/STTR program. Thus, this team of people would form an X-Lab, where the key investment (say $50M/yr for 3-5 years, in a milestone-driven contracting method) would come from NSF/TIP. This is not a priori crazy - multiple other groups have looked at non-profit startups as a way to fund science. The program was announced at a level of $150M/yr for ten years. (The Science article implies that those in charge want a lot more money now than was in the TIPS appropriation plan for this year. Here is a claim that this is not true, which would make the cuts even harder to understand.)
One big catch: The way the X-Labs are being implemented seems pretty inflexible. An X-Lab has to be its own entirely independent ("autonomous") entity (rather like a company or non-profit), not a subsidiary or an operating unit of a company or university. Any senior personnel involved are required to be 100% full-time associated with the X-Lab. That means that anyone doing this from a company or national lab would have to quit their previous job or go on a complete leave of some kind. Anyone doing this from a university would have to resign their faculty position or go on a complete leave of some kind. Issues like IP and benefits/health insurance seem nontrivial and not worked out. Given the current uncertainties with everything associated with the NSF, this is quite a proposition for established researchers to undertake.
So, here we are, with reporting that there will be large cuts across the NSF, regardless of what the appropriations said. Anyone with first-hand knowledge who wants to chime in, please weigh in in the comments, or drop me a line (presumably from a non-NSF email address).
As bad as this is, the part of the article that truly angered me was this:
Program managers would normally rush to inform potential and current grantees about such dramatic changes. But the memo tells program managers to keep their mouths shut. “This information is highly confidential,” it reads. “Please do not communicate anything to PIs [principal investigators].”
Really?
You know this is not supported by the actual program officers, because this "highly confidential" information was almost immediately sent to a reporter. Daylight is a great disinfectant. Public pressure and Congressional pushing forced NSF leadership to relent on the plan to destroy the Ocean Observatories Initiative. Maybe making this budget cutting known can focus attention on this, rather than having drastically reduced NSF research funding be a fait accompli.
This is an utter disaster for science in the US. The goal seems to mainly fund AI at the expense of everything else, but with some Quantum on the side.
ReplyDeleteQuantum is a perfect example however of how blue-sky, fundamental physics and mathematics research played a key role in creating and transforming what is now a multi-billion dollar industry.
Two primary technologies in the push towards scalable quantum computers rely on topological ideas that trace their intellectual origins back to string theory. These are topological quantum computation with Ising anyons (Microsoft) and topological quantum error correction using surface code (Google superconducting Josephson junctions, trapped ion efforts like Semeghini and Lukin at Harvard). Both threads are largely due to Kitaev, but Kitaev's work draws heavily on ideas of conformal field theory.
Anyons, braiding, and fusion rules that underpin both technologies were originally developed in the context of classifying rational conformal field theories by Moore and Seiberg. CFT in turn was originally developed as a platform for solving the 4D string by BPZ. This effort failed (as did the hope that classifying rational CFTs would produce a string theory matching our world). However, as a byproduct BPZ produced the formalism to solve 2D classical or 1+1-D quantum critical phenomena. Witten eventually linked topological field theory at the heart of modern QC efforts to CFT, demonstrating that ground state wave functions and 1+1-D edge states are just different quantization schemes for the same 2+1-D topological field theory.
String theory itself was a failed attempt to solve the strong interaction. Regardless of whether SUSY or higher dimensions actually describe some version of our (seemingly de Sitter) universe, these mostly pure intellectual musings eventually lead through the efforts of many, many academic scientists (not just the big names above) to the technological push we have today.
Cutting the division of mathematical and physical sciences to 1970s levels in favor of million-dollar AI start-ups will very likely backfire spectacularly, and the US will pay the price in terms of innovation on the global stage in the long run.
1. Many of the aspects of the toric code were understood before Kitaev, in the context of resonating valence bond states, but he provided an exactly solved model. He acknowledges this himself at the beginning of chapter 3 of these notes (https://arxiv.org/abs/0904.2771).
Delete2. Anyons, braid groups and topological phase transitions were initially thought about in the context of generalized Ising models in the 1970s and even the late 1960s. The first use of anyons was in explaining the FQHE (https://arxiv.org/abs/2212.12632)
3. Even before BPZ, P was thinking about the conformal symmetry of critical fluctuations, so the precedent is again in condensed matter (A. M. Polyakov, “Conformal symmetry of critical fluctuations,” Sov. Phys. JETP Lett. 12 (1970) 381). Polyakov seems to have moved between both worlds.
There is a connection to string theory, but the more relevant precedents for these things are in condensed matter.
I agree that blue-sky research is quite valuable and necessary, and these are examples that show that. Don't overhype string theory, though. There's enough of that already.
Agreed there were many condensed matter precedents, and probably I was too fast and loose with Anyons. But fusion and braiding (i.e. "modular tensor category theory") was introduced to CFT in a very formal way for ends that had nothing to do with CM, and yet conformal blocks eventually were linked to topological order and ground-state wave functions. Many formal aspects of topological field theory also have stringy origins, i.e. the Verlinde formula. Cardy brought modular invariance into CMT and statistical physics, but I suspect that wouldn't have happened nearly so early without his particular high-energy background.
DeleteDon't want to get into the "string wars" except to say that indeed ST was oversold to the public (given the landscape problem emerged in the 80s).
At the same time, in CMT there has long been a bias, sometimes subtle and sometimes not, against acknowledging the role of blue-sky theory. The preference is towards theory that immediately explains existing experiments, and while that is understandable, there is no guarantee that the intellectual framework necessary to understand any particular physical phenomenon exists at the time it is discovered. What seems to be serendipity is often due to the decades of obscure theory building before an application springs into existence.
Now, with the push for AI and "automated workflows" I fear that theoretical science driven by creativity and intuition will be fully sidelined when it comes to grants (and potentially journals as well).
Re: Now, with the push for AI and "automated workflows" I fear that theoretical science driven by creativity and intuition will be fully sidelined when it comes to grants (and potentially journals as well).
DeleteOn the other hand, AI is great at exactly using existing ideas from "far away" to apply them to practical problems at hand.
So most that are solvable with existing ideas (even if we don't know that yet)may be solved soon. And that may lead to more proper blue sky, serendipity, and foundational thinking for those that are no longer distracted by finding mundane solutions.
(Have to stay optimistic somehow these days...)
This does seem like dire NSF funded activities and probably is hurting US research excellence ... truly gives China leverage without having to change their behavior. I'm curious what the DOE equivalent graphs look like? Is this across the board included DOD (I guess its DOW now 😵💫)? Whats the feeling of program managers? Are they thinking AI solves many research issues for us so why over fund?
ReplyDeleteEven the comparisons to 1973 are misleading as there are many many more grant applications and scientists now, so an equivalent # of successful grants means much lower funding odds.
ReplyDeleteRegarding the historical funding trend and how today's budget and competition level compare to 1973, I think that in addition to the much higher number of university PIs eligible to apply, the single-investigator award size is now nearly five to ten times what it was back then due to inflation (and perhaps also to increases in overhead/indirect cost rates over the years, as state funding for most universities has shrunk). So the competition has become rather extreme at the current NSF funding levels. Compared to the 1970s, I'd describe today's level of competition as essentially a random selection process among the proposals that panels rate as competitive or highly competitive.
ReplyDeletehttps://www.aps.org/initiatives/advocate-amplify/policy/federal-grants-rule-change
ReplyDeleteFolks, read the room. You’re supposed to be joining me in doom-spiraling about the public research ecosystem, not doom-spiraling about AI. (Kidding.). NSF has put the draft update of the proposal guide up for public comment btw. Honestly, at this point I’m surprised OMB isn’t saying agencies can only accept proposals written in Linear A or the mysterious pseudolanguage of the Voynich Manuscript. https://www.regulations.gov/document/NSF-2026-OTR-0001-0003
ReplyDeleteHence the aps effort url
Delete