My Lords, I am very pleased to follow the noble Lord, Lord Ravensdale. Two of the amendments in this first group are in my name, Amendments 25 and 27, and I want to speak to Amendment 27 first. It is grouped with Amendment 1 because we start by debating, quite properly, the purposes of ARIA as an agency. What is it here to achieve?
As the noble Lord, Lord Ravensdale, said, we are not seeking to replicate DARPA but to learn from it. DARPA said that its sense of mission was part of the reason for its success. However, that mission in this context was originally
“to prevent and create technological surprise”.
That is an interesting concept—to prevent technological surprise happening to the American Government and, at the same time, to create technological surprise on its own part. One might say that you could substitute “create technological advantage” in the latter case. Interestingly, in more recent years, when DARPA staff were asked what they regarded as their mission, they said it was to be part of “shaping the future”. Indeed, I think that is where our starting point should be. We want ARIA as an agency to be part of shaping the future.
My problem with Amendment 1—actually, I do not have a problem with Amendment 1, because you could stretch the language of sustainability anywhere; that is its advantage but also its problem. I am not sure I understand what the board of ARIA, or its leading members, would interpret as being outside the scope of the sustainability criterion. Does it actually help them? I am not sure that it does. If anything, they might feel that it constrains them towards certain missions. The DARPA example we ought to learn from is that, in practice, it set out to define for itself a range of missions within the organisation.
I note that sitting next to the noble Lord, Lord Ravensdale, is the noble Lord, Lord Broers. I take from his Second Reading speech the thought that the programme managers are at the heart of this system, and the programme managers are chosen in relation to the programmes that DARPA is pursuing. I suspect the same will have to be true for ARIA—that it has to decide, “What are our programmes?” The programmes, in my view, might be mission-led—for example, related to adaptation to climate change—but at the same time they might be technology-led. For example, they might be to pursue AI and the data economy or to look at cell or gene therapy. There is a range of those possibilities. We need to give ARIA, as an organisation, the flexibility to decide the missions that it thinks fulfils its purposes. The missions will develop over time, but the legislation cannot change repeatedly over time, so the legislation should be sufficient to enable ARIA to select the missions it wants for the future.
My Amendment 27 is in this group. A report of July 2016 produced for DARPA about innovation in DARPA isolated four “sources of success”, as it put it, the first of which was the “limited tenure” of the leading executive members
“and the urgency it promotes”—
nobody was appointed for a period exceeding five years. The second was a “sense of mission”, which I was just talking about. The third was “Trust and autonomy”—both giving DARPA autonomy but also within the organisation trusting and giving autonomy to the programme managers in particular. The fourth was:
“Risk-taking and tolerance of failure”,
which of course we are setting out to incorporate into this legislation for ARIA. I add that DARPA interpreted this as meaning “Move fast and take risks”—do not spend a great deal of time trying to assess all the risks, because you could lose the opportunities in the process.
Amendment 27 seeks to replace the language of Clause 3, not because I have any objection to the purposes set out in Clause 3; my objection is to the drafting. It says:
“ARIA may give particular weight”—
I am afraid I do not understand what is meant by “particular weight” or how people who read it subsequently will know what that means in this context—
“to the potential for significant benefits”.
We are all agreed about “significant benefits” and we know what they are because they are in Clause 2(6) above. It then refers to
“research … that carries a high risk of failure.”
It is awfully close to being a piece of legislation that says that ARIA should look for projects that are quite likely to fail because those are likely to give the most significant benefits.
This is not the approach that legislation should take. Legislation should be more deliberate. I thought: what are Ministers actually looking to do in this clause? I agree with the noble Lord, Lord Ravensdale, in his Amendment 1. Ministers—and we—are looking for ARIA to seek to have transformational effects. I think we are agreed about that. That is why Amendment 27 refers to “transformational effects”. I have also included a reference to the possibility of technological advance through
“the development and exploitation of … research”.
I do not think that is mentioned elsewhere but I think it is helpful because, actually, many of the advances that have occurred, including in DARPA’s programmes, were not themselves the object of the mission but were the result of the process of discovery and curiosity and the exploitation of research.
4.30 pm
Turning to the question of benefits, where the realisation of benefits is uncertain but is expected to be significant and “significantly to exceed costs”—although that may not be necessary—ARIA must accept, under Amendment 27,
“that the projects it supports may entail a high risk of failure.”
It is a very simple proposition whose language is plainer than the original drafting. ARIA must look for significant benefits, transformational effects and understand the potential for technological advance through the exploitation of research, but accept that where significant benefits can be achieved, there may be a high risk of failure.
This drafting is longer, but by being so it is also a bit more sequential regarding the way in which ARIA would be asked to do its job. However, it is not trying to tell ARIA which technologies to use, which missions to undertake and what precisely it should set out to do. If we want to do that, we have a shedload of it in UK Research and Innovation. There are institutes with specific objectives and challenge funds with specific objectives. There is the Industrial Strategy Challenge Fund, and “challenges” for the ageing society, clean growth, AI and the data economy. There are many ways in which we can set out using our research and innovation resources to try to deliver specific objectives. In my view, ARIA is not about that; it is about trying to achieve transformational effects by thinking outside that framework. I do not want us, through this Bill, to put into legislation a framework that inhibits ARIA’s ability to think laterally, horizontally, outside the box: to think about technology in ways that are different from the rest of the UK Research and Innovation landscape.
Finally, Amendment 25 is a probing amendment for precisely that purpose: what is the point of ARIA if it duplicates what UK Research and Innovation is doing? We should therefore make it clear to ARIA that it should at any given time take into account the strategy of UKRI. If it does not, the risk of duplication is high. ARIA, by its nature, should be seeking to do something which UKRI is not.
I hope that my noble friend the Minister will see some merit in both these amendments, even if we might disagree at this stage about the drafting.