White House report says Trump can usher in a "new golden age" of science

The report also seems not to realize that it is being written by the Trump administration. One of the reasons researchers have spent so much time writing grants is a general uncertainty about funding, and that’s something that the administration has dramatically increased through the haphazard termination of existing grants, by proposing massive budget cuts for science agencies, and by slowing the release of funds that had been allocated to researchers by Congress. This lack of self-awareness permeates the document, but it’s worth looking at a second example. In several instances, the report highlights the Human Genome Project as an example of the sort of project that government excels at driving. But it seems to be unaware that the project was a major international collaboration, with entire chromosomes being sequenced outside the US. Pride in the completion of the genome seems out of place from an administration that is actively trying to minimize international collaborations involving the researchers it funds. In the same way, it highlights how the Human Genome Project required the immediate and open release of the data it generated as an example of what the Trump administration considers “gold standard science.” But SNGA also decries how making results accessible to all has allowed other countries to develop industries based on advances that occurred within the US. In short, the report lacks a perspective that’s internally coherent, or consistent with the other goals and actions of the Trump administration. Yet the SNGA treats this as settled and uncontroversial. It talks a lot about how we need to promote more innovative science, and suggests a wide range of different funding models that might do so. It also says agencies will need to evaluate whether the grants they are funding are accomplishing what was intend. But it never describes how agencies should objectively measure innovativeness, so it’s unclear how they can perform that evaluation, or whether the scientific community will consider their results valid. On the flipside, this is one of the cases where SNGA’s proposed solutions align with Trump administration actions, specifically its call to reduce the importance of peer review. In the new document, peer review is presented as part of the problem: “Review panels often gatekeep proposals by consensus, disincentivizing transformative ideas.” And many of the new funding mechanisms it proposes include reviews by single individuals who may or may not be trained scientists. While this is likely to broaden the scope of ideas that get funded, it also seems likely that it will increase the funding of fringe ideas, something that the OSTP does not seem to consider. (I’ll note that this is consistent with Trump administration actions and not ideas; peer review remains part of what the administration is calling gold standard science, which SNGA also endorses.) Aside from innovations within science, SNGA also is concerned with the innovation that occurs when science gets translated into products and processes by commercial interests. Here, the report decries what it terms a linear model, one where government funds basic science, which the market then takes on a one way journey to commercialization. Again, this is a case where people in science would likely agree that things are often considerably more complex. The report favors a model where commercialization is more of a conversation, as technology developments allow new scientific work that in turn leads to new or enhanced opportunities for further commercialization. There have definitely been instances of this. A great example is the real-time RT-PCR tests that were initially used during the pandemic. Commercialization of both reverse transcriptase (the RT) and PCR allowed scientists to develop the real-time monitoring of reaction progress for their own research. Companies then commercialized hardware that simplified the process, and still other companies then developed diagnostic tests using it. But SNGA puts all of its eggs in that basket, which is just as incomplete as the linear model. Quantum mechanics stayed trapped in physics departments for roughly 50 years before it got commercialized via lasers and semiconductors. And it’s hard to imagine the timeline in which we’ll end up commercializing something like the detection of gravitational waves. Neither fit the linear or conversation models of innovation. The whole idea of government funding for basic science was the recognition that it’s not possible to predict in advance which scientific findings will have commercial applications, so companies, for the most part, weren’t going to do it. While SNGA is right that we should encourage those cycles of science-technology innovation when we find them, it offers no suggestions for how we identify one before it takes off and government funding becomes irrelevant. Because that is the only type of technology development it acknowledges exists, it has little to offer for any others. Here, its solution is grand in scope. It views the maker community, which is mostly interested in scratching personal itches, as a sign that the US public wants to make things again. The SNGA authors figure that they would be perfectly happy making things for science. “Establish national fellowships for skilled craftspeople,” they suggest, “practitioner-in-residence programs embedding machinists and technicians alongside Ph.D. researchers, and portable industry-recognized credentials in advanced manufacturing and lab techniques.” It’s unclear what these machinists will be doing with the researchers. But the results will be glorious: “By rebuilding the link between science and hands-on craft, federal leadership can ensure that the economic returns of discovery, including the jobs, supplier networks, and process knowledge encoded in the hands of workers, accrue to Americans.” Its example is Detroit in its heyday, although that seems to have been a period of engineering refinements that lacked a notable scientific component. And it’s unclear what scientific research center the SNGA associates with Detroit at that time. Are there some fields that can be revolutionized by things like a 3D printer and a bit of time with the maker community? Robotics seems like an obvious choice; developmental biology does not. Elsewhere, the report returns to the familiar claim that regulations are also stifling the potential for science-driven commercial innovation. The only specific area that’s cited, however, is nuclear power: “nuclear energy stalled in America not because the physics failed, but because regulatory choices over the past half-century made building uneconomical.” That is not a realistic reading of the history, as there is little indication that regulations are the primary cause of the massive delays and cost overruns that plague nuclear plant construction. The complaints about regulations are one of the many cases where SNGA descends into political and cultural grievance. The most obvious case is where it detours to grumble about the role of the scientific community in the arguments over COVID school closures, but there are many additional ones. It is once again dismissive of diversity, equity, and inclusion (DEI), while saying that all Americans with aptitude need to have access to scientific training—exactly what DEI programs were meant to ensure. It also complains that foreign students were taking slots in PhD programs from deserving Americans, while ignoring the reality that the administration’s attacks on science funding have caused a number of schools to cut the number of students they admit. The fact that those foreign students often want to stay in the US to contribute to either scientific or commercial endeavors is ignored. Grant overheads allow the institutions that host federally funded research to pay for the upkeep of the facilities where science happens. But they’ve been targeted by the administration, so the SNGA takes time to complain about them as well. Most strikingly, the administration that terminated grants wholesale due to political disagreements with the subjects they were funding had the audacity to argue that it would lead the charge to “ensure that selection [of grants] rests purely on merit, not the political fashions of the day.” Beyond the political grievances, the report seems to have been shaped by OSTP head Kratsios’ time as a venture capitalist. Venture capital is presented as a separate source of scientific funding from government and commercial. SNGA presents it as potentially more rigorous, since companies will fail if they’re based on faulty scientific ideas. Left out is a consideration of the fact that most venture-backed companies do in fact fold; this is a track record that would be considered problematic by the authors of this report if it involved grant-backed projects. This attitude pervades the report’s approach to AI, which largely buys into the biggest hype imaginable. The reality is that we’re still in the process of understanding what types of AI developments will have an impact in which fields. But SNGA envisions a future that, well, deserves to be read in full: These pieces lay the foundation for a continuous, market-mediated, agent-based scientific economy. Imagine a funder posting a million-dollar bounty for the first validated therapeutic target for a rare disease. An agent working on adjacent problems notices a promising lead and posts a smaller bounty for replicating the finding. Other agents assess whether the problem falls within their competence, bid for the work, and contract an autonomous laboratory accessible through the internet, which runs the experiment and returns cryptographically signed results. But it’s not clear what dream the people who wrote the new report are pursuing. The Endless Frontier was a serious effort to make the case that it was in the US’s national interest to fund basic science, even if there wasn’t a clear commercial or national security endpoint to the work. Its audience was the policymakers that could turn that into a reality. A New Golden Age, in contrast, is a grab bag of political grievances, half-thought-through justifications for policies the Trump administration was already pursuing, and insinuations that scientific experts are kind of annoying and should learn how to fix car engines. It’s unclear who the report’s audience is. Certainly not the policymakers in the current administration, since it repeatedly indicates that the decisions they are making are great already. And it’s certainly not the scientists, who are currently suffering from the impact of those decisions. A similar ambiguity exists about the goals of the document. There are lots of specific suggestions: get new funding mechanisms; don’t pay attention to peer review; hire some mechanics; get industry more involved; support AI, as it will change everything. But if those added up to a coherent whole, I was not able to identify it. “A decade from now, American researchers should look back at our work and say: ‘The vital questions I could not pursue then, I am free to pursue now,’” Kratsios wrote about the document, in what seems to be the clearest indication of a goal. But A New Golden Age provides no reason to expect that they will.