June 20, 2021

Can't Buy Me Innovation

A few months back, the Government announced its plans for some big, grand capital-I innovation programme that they'd like to invent the next Internet or something.

It's the sort of thing that in theory I'd be very interested in, but I've learned to mostly ignore such developments. I'm pretty confident that it'll achieve little beyond wasting huge amounts of public money. That's a genuine shame, but I think the best way to deal with that is to build an alternative approach that makes it obsolete. So that's where I'm focusing my efforts.

However, since Laura James pointed at this excellent blog post on ARIA (the name of the Government's initiative) from Nick Hunn, and Rachel Coldicutt live-tweeted a thread about Dominic Cummings' appearance at the Science and Technology Committee talking about ARIA, some thoughts about it have been rattling round my head. Let's see if I can get them out and into this post.

Nick's blog post does an excellent job of explaining the cultural problems with the UK R&D funding landscape, which won't be solved by throwing more money at it:

The people who put together the scoring schemes for our current development grants would run screaming from the room at the prospect of a mere 20% success rate, as would almost every civil servant. This is despite the fact that the real success rate of current grants, at least in terms of game-changing innovation, barely registers. The people administering them have invented some very novel scoring criteria which make them look as if they are successful, but that is probably the most innovative thing which has come out of them.

I think part of the problem is that they don't know what they want to "buy", beyond "innovation". And the problem with that is that it isn't something you can buy. When asked what they did today nobody replies "I innovated", it's "I wondered what would happen if..." or "I thought it would be interesting to try...".

Those doing the work, and potentially innovating, are too close to the work to make the call. Is anything I do innovation? Either it almost all is, or hardly any.

The ur-example of this sort of approach is always DARPA. The difference there is that they weren't buying "innovation", they were buying "new defence technology". That's actual things rather than a vague concept. I think that's a key improvement that could be made to the UK approach: not defence spending, but spending on things that Government wants or needs. That will mean that:

  • You'll have an idea of if you succeeded or not. Failures are okay, and this way at least you'll know if you've had any!
  • Some level of commercialisation is likely to have taken place.

Maybe I'm just arguing for the cash to be given to the SBRI (the Small Business Research Initiative, which funds R&D from small businesses responding to challenges from Government). That'd be more useful than many of the other options. The problem then becomes one of generating the ambitious, interesting ideas of what to buy, which is likely to be just as susceptible to capture by those good at networking and navigating bureaucracy. We'd surface more of the failures though, which you'd hope would provide more of a corrective feedback loop than the present system.

Moving onto Dominic Cummings, lots of discussion has been about his love of "weirdos and misfits". While I agree that he'd be a poor arbiter of weirdos and misfits, I think that a diverse and alternative group of misfits and weirdos would be a good way to increase the variety in areas explored.

Cummings mentions Bell Labs, and as Making Art Work documents, that did enable a bunch of interesting art and tech crossovers.

One of the replies to Rachel's thread makes the important point that "weirdos and misfits" mustn't be conflated with the myth of the lone genius. Definitely. This is a place for Brian Eno's scenius.

Or for a different collective approach, we could look to The Lucas Plan. They got so close in the 1970s(!) to pursuing a bunch of technologies that we could really have benefited from now; think what they could have achieved with backing and funding.

If you find the right sort of weirdos and misfits then the hardest part is going to be persuading them to take the funding. Anyone doing something interesting and different is going to have encountered, and been failed by, the existing system. The independence to follow your own interests, hunches and research is not given up lightly.

The DoES Liverpool community is a good example of the sort of scenius and cross-pollination of ideas between misfits and weirdos, with interests in IoT, plastics recycling, CNC tools, knitting, civic software, biomaterials, and more, plus permutations and combinations of all of them. With no funding in its decade of operating, and despite a much broader remit than just innovation, it runs rings around other innovation hubs funded to the tune of double-digit millions.

That's not to say that we couldn't do more if we were generously funded. We could have more space, and more equipment, and members of the community could be paid to follow their interests full time rather than having to balance that with other work to pay the bills. But it's the people, not the money, that's the important differentiator.

Another of the replies in Rachel's thread summed it up nicely: "the really radical answer is a living wage UBI". Let people follow their interests, that will unlock more "innovation" than any Government quango (or private corporation, for that matter).

Posted by Adrian at June 20, 2021 10:40 PM | TrackBack

This blog post is on the personal blog of Adrian McEwen. If you want to explore the site a bit further, it might be worth having a look at the most recent entries or look through the archives or categories over on the left.

You can receive updates whenever a new post is written by subscribing to the recent posts RSS feed or

Post a comment

Remember personal info?

Note: I'm running the MT-Keystrokes plugin to filter out spam comments, which unfortunately means you have to have Javascript turned on to be able to comment.