A case for British DARPA
Dominic Cummings should probably have been put in an institution akin to Arkham Asylum instead of being an advisor to Number 10, but he was absolutely right about having a British DARPA. With him gone it probably won’t happen, and this is a tragic mistake.
Often (like, once a week lately), I end up in policy discussions about the lack of [deep tech, scaleups, European Silicon Valley], and while there are pages of zero-cost policy interventions that could be enacted before helicoptering money (I have a long list if you want) I think the most important thing is to finally accept that the EU R&D budgets have been, to put it mildly, underperforming.
I have been on the receiving end of those for 25 years (barely, as I think the entirety of grants my companies have received has been <$50k vs millions in VC & FDI) and I have always been impressed by the entire shitshow which that process is. There was never a link to outcomes, or being the causal factor of the R&D effort. Instead it was some sort of matching money (after the fact, but one can’t start until approved 🙄), with lots of nice paperwork that nobody was ever going to read or understand. Pages and pages of fluff, each identical to each other. All empty calories. All created by this underbelly of R&D advisors that just add more inefficiency to the system. And the result is companies like Amazon that single-handedly outspend the entire EU R&D budget (and Europe doesn’t have anything near a homegrown sovereign cloud infrastructure). Then there are Google/FB/MSFT/Apple too. And then add their Chinese equivalents.
The largest free market in the world lags behind when it comes to technology.
The UK until now was in the same bind because it was part of the EU framework - which to some extent limited funding structures. And yet, countries like Finland are outspent 10x in R&D by Germany, but because of their different choices in capital allocation, Finland generated some of the largest companies in Europe, as well as on a per capita basis hitting it out of the park. The UK in the meantime has ARM (founded thanks to American capital, currently owned by a Saudi-Japanese consortium, soon to be acquired by a US company), which was worth 5% of Intel’s, and did as many acquisitions in the last 20 years as Cisco, an also-ran, does in one bad year.
Innovate UK’s mandate is too wide (and still underfunded given its ambition). The MoD notoriously has had their budgets under lockdown, and even minor expenses are scrutinized, so that bleeding-edge R&D happens elsewhere. Defence budgets have been decreasing, research privatized (creating QinetiQ, which at $2.1bn is a nice outcome but to put it context it’s also 2x in value as the not-ready-meal startup Gousto, and various other spinouts, none of which reached scale), and a transition to a software/AI/cybersecurity/drone type of platform make a challenging backdrop.
The latest expansion of defence spending won’t really go towards new stuff, but mostly to maintain our dwindling capabilities.
So I don’t want to go all Mazzuccato on you, but the only outcome would be something exactly like DARPA; a separate entity under the MoD (and not UK R&I, but certainly have a liaison officer assigned), staffed by technical experts, focused on outcomes of long term research, with minimal bureaucracy (if you want to see this in action, look at Fast Grants), and with a budget growing to £500m/year in 10 years. It is a monstrous amount of money, especially for a country where 2m people use food banks and poverty is normalized. But spending more on welfare only helps today, and not tomorrow.
This would enable the UK to become the primary exporter of technology (as opposed to being the primary exporter of science which is today), and have a larger number of scaleups in Europe in 2040.
The defence of the realm is a priority and it is not just about borders, it is about technological hegemony and economic supremacy. And it’s nothing to be ashamed about.
One last note, when I said “like DARPA” I mean warts and all. The first few years of DARPA were a massive clusterfuck, and not just the “money wasted” type, more like the “people killed” type. It took a while for DARPA (and its predecessors like ARPA) to find their niche, but the result is being instrumental in the creation of the basis technology for EVERYTHING we use on a daily basis. We live in DARPA’s world. It is time for the UK to build its own version of it.
It will certainly require Parliamentary approval, but in order for that to happen MPs and their constituents need to be convinced that this is the way forward, and that a pound spent on it will be a pound not spent elsewhere. And that it is worth it.
 DARPA currenly employs ~200 people, Innovate UK has 250, and its main body UK Research and Innovation employs a staggering 7,500 people
 Conveniently this is similar to DARPA’s budget which is ~$10 per capita