LAST month the Advanced Research and Invention Agency (Aria) announced the appointment of a new CEO, in line with its intention to launch around the end of this year.

Aria is a new funding agency modelled on venture capital funding for high-risk “moonshot” research that will sit alongside, although independent of the umbrella British research funding agency UK Research and Innovation.

The new CEO, Ilan Gur, has a background in venture capital.  

Aria is closely associated with Dominic Cummings. When in government he pushed his obsession with disruptive technocracy.

He was frustrated with traditionalism and gatekeeping in science research, lionising instead Silicon Valley and frequent failure of start-up companies.

Aria’s budget is £800 million over four years — around double the total funding for arts and humanities research.

Though this is still small compared to the £7.8 billion a year of overall science funding through existing research councils, Aria will make funding decisions faster and based on less evidence.

The intended outcome of this model is that funding can be directed to unknown and underfunded researchers without the backing of major Establishment voices, and give them freedom to pursue their ideas singlemindedly.

The model is the US defence funder Darpa (the Defense Advanced Research Projects Agency), which is widely credited with visionary funding in the 1960s that paved the way for future technologies like the internet.

Aria hopes to fulfil the venture capitalist sci-fi fantasy of being responsible for “the next internet.”

Currently the funding landscape for research in Britain is dominated by the largest research-intensive universities.

Research proposals are funded after time-consuming review by committees of experts in the field, often with a high degree of vested interest in the future direction of the field.

The process of writing and submitting these documents is labour intensive; most submitted research proposals are not funded.

Proposals place the strongest emphasis on the clear and obvious outcomes, in fact the inevitability, of the proposed research.

No matter how novel and original the idea the researcher might want to follow, the success of the research proposal is often dependent on experts conservatively evaluating their existing expertise.

This system feeds off paternalism and networks of influence, particularly in the highly hierarchical context of academia.

Aria, it is claimed, will change this by funding outsiders with crazy unproven ideas. The aim is to increase the appetite for risk and experimentation.

Like venture capitalists, by funding enough failures they hope to produce a small number of huge successes. Is it likely to achieve this aim? We admit it’s an exciting prospect.

This is because the current state of funding sets a low bar. Research worth its name should involve a risk that it might fail. That’s not a currently accepted framework. This means that all research proposals promise conservative and supposedly risk-free investment of public money.

The only rewarded research outcome is that research was successful and meaningful, so scientists are pushed into claiming that everything they try works out.

In this respect, the principles behind Aria differ from how the Conservatives have previously approached research funding.

Universities have long been portrayed as wasting money on frivolous and theoretical projects by the right-wing press.

As a result, funding bodies have been encouraged to adopt a timid and value-for-money approach to research investment, emphasising spin-out value.

In general, the reaction of scientists to Aria has been the same as always for new funding announcements: cautious optimism about money for science, but wanting to see it work in practice.

This encapsulates an inherent conservatism. The allegiance to an imperfect status quo often involves a need to believe that current science funding is meritocratic.

The huge amounts of effort required to put together many grant proposals means that only large and already successful research teams can apply. Capital attracts more capital. Those who begin to fail quickly fall by the wayside.

Research on 1.2 million PhD theses found that those written by people from demographically under-represented groups tend to contain more original ideas, but those people go on to be less successful in science. Defending the existing bureaucracy does not mean defending fairness.

Much of the criticism so far from scientists of Aria has revolved around its exemption from the Freedom of Information (FOI) Act. As Nature observed, this puts it in the vaunted company of the royal family and the security agencies.

It is an easy criticism; the funding of research from public money should be transparent.

This suggests that Aria may focus on projects in “defence” (or more accurately, “attack”). But the government doesn’t struggle now to fund university research on these morally reprehensible topics through routine processes that are subject to FOI. 

More likely, it seems that the reason for dodging scrutiny is that Aria’s processes will be capricious by construction. One of the purposes of the agency is to cut through “red tape,” which is why it fits with current Conservative ideology.

The hierarchy of Aria places enormous faith and trust in a small number of people. Theoretically, a crazy idea scrawled on a napkin might suffice to get public money — if senior Aria staff like the cut of your jib.

Giving Aria this power and calling for unconventional ideas is bound to result in not just productive failure but avoidable disasters.

Snake-oil salespeople will flock to it. Venture capitalism is plagued by biomedical fraudsters like Elizabeth Holmes, the founder of Theranos.

It’s likely that by exempting Aria from FOI, the government wants to avoid embarrassment similar to the successful legal challenges to procurement processes during Covid-19.

A venture capital mindset with public money could possibly produce big rewards, but the “investors” deciding who gets money are not using their own money. Their own job is also probably not at risk, so the risk-reward equation is different.

The model is incomplete. Removing FOI could mean the risk to investors is reduced even further — it may not produce the right results.

The government’s statement setting out the rationale for Aria says that it will use a “programme manager” funding model, giving individual scientists a lot of freedom.

It gives a specific example: an industrial engineer with “a vision for a flying personal transport device, with potential commercial, military, and/or home security applications.”

The engineer is quickly given £50,000 to flesh out their application, with a successful application followed by a much larger budget of £20m to be a programme manager over five years.

The example highlights how only need two approvals are needed for each decision, so the engineer “can start and stop new projects at will as new evidence emerges.”

Over the course of the five years they commission research from leading universities and forge “a community of interested parties around an emerging technology” in order to try and turn their idea into a commercially viable product.

The example is interesting — because of how boring it is. Drones already exist. And venture capitalists already fund ideas with obvious commercial applications to the tune of millions of pounds.

Aria funding seems to be nothing more than an actual venture capitalist fund, run by the government. If Aria is committed to the most high-risk projects that produce transformational change, then they will have little obvious applications at the beginning.

Any requirements for clear commercial application will only fund the exact same commercial projects that are already well-served. This isn’t a route to funding blue-sky ideas.

The idea that the project manager will see their vision into reality by commissioning research from top universities is also unimaginative. It means that Aria funding will just accumulate at the universities that already acquire the most funding through existing routes.

Even though Aria itself may be an “agile” funding body with little red tape, those they fund will still need to navigate the university bureaucracy that can delay current research.

The proposal that some science funding might be earmarked to give higher-risk ideas a chance is exciting. Blue-sky research can lead to true transformations. More often, it fails.

This is the reality of science, so we welcome the proposal to normalise the failure of scientific research as one of the costs of new knowledge. But this work-in-progress research agency falls short of this bold potential.

The case study suggests it will end up as an unoriginal attempt to replicate current venture capitalism — competing directly with existing venture capital formulas. What is the point?