Read:

How should we fund scientific research in the UK? Is the proposed Advanced Research and Invention Agency a good idea?

You could argue that the current system is incredibly successful in that we do have this amazing science base. There is a slight danger of don't break what's working. But the reality is we spend a very large amount of money on scientific research in the UK, which is great. But what we haven't done is really look at impact. So, one of the things that I was advising the Prime Minister on - I was very forceful in making one of the requirements now for getting a grant, at the end of it you have to say what would the impact of this idea be, that wasn't even there before. It might be that it's got no impact at all because its blue skies research, which is fine, but you at least have to think about it for 15 minutes. So, I think there are a series of changes.

The UKRI, which is the funding body, is a very large bureaucratic body. Some people refer to it as a blob. And the problem with it is , because it's all peer based, it has some fundamental issues. For example, it's very hard to kill something. If you imagine that if you're a scientist working on whether snails ever flew in the past, then you get some grant money and get some PhD students and before you know it, you have a few departments around the country that specialised in the evolution of flying snails. Well the problem is whenever a grant request comes in, the only people you can talk to are the experts on the evolution of flying snails, who of course, will approve it. Whereas actually at some point someone needs to say: no there's no evidence snails ever flew, we’re killing that one. And so there are things that need to be fixed there.

The other part of the government's approaches is Innovate UK. I think Innovate UK needs a massive overhaul, it's actually damaging the UK, it creates a grant culture. So, if you have a little idea and you're spinning it out in university, you get a grant. The problem is you get another grant, and you get another grant. A simple way of fixing that would be to say, you get two grants, the third one you need to show me the product, the fourth one you need to show me some sales. And then that would force the commercialization rather than what you've had is people surviving on the grant culture in the UK.

For a scientist, it's uncomfortable learning about how a salesforce works but if you want to make a company you’ve got to do that. So, the question is, do we need to cut through all this? I think we do. And obviously the model for that could be DARPA in the US, it's this ARIA idea. It will upset the applecart, there will be lots of the blob that will want to kill it. But it is really important because it just needs to set out, without the bureaucracy, and do things because the modern world is moving so fast that you need to have something that can move fast, it can make bets that are not necessarily totally justifiable, and they can try and hit some jackpots. And so, I'm a great supporter of the idea, I think it does need to be defended though because, as is often the case with such initiatives, the blob comes along and sort of subsumes it and it disappears inside. If we can keep it separate and we can keep it on mission and it's driven in a very strong and decisive way, it could bring amazing things because it can draw on all the amazing things inside the UK research infrastructure, and then start trying to turn that into some amazing things very quickly.