Prior to joining TU Delft, Scott Cunningham worked in the computer and software industry, creating analytical models for commercial clients. His work on national innovation indicators helps inform policy for the governments of the U.S., the U.K. and Malaysia. Scott is interested in operations research and decision sciences approaches for policy making. In particular he is interested in probabilistic models of social exchange. Other interests include building multi-actor systems theory through the economic sociology and innovation policy literatures. A recent publication is Tech Mining (with Alan Porter), a book on assessing new technology developments.

Can you envisage any major wild cards, positive or negative that may occur in the next 20 years?

Two very positive wild cards. The first is the idea that we might be on the brink of a digital renaissance. Some commentators have suggested that we are facing a lot of intellectual surplus and this is being harvested and utilized by the internet and new institutions surrounding the internet. I am also interested in the topic of material plenty in Europe. There is a lot of worry in Europe about basic necessities in the rest of the world. But there is also an interesting scenario where Europe may have all its basic needs for energy, transport, food and manufacturing met, freeing more resources for creativity and intellectual enterprise.

What do you see would be the most dramatic impact of these two wild cards?

I have looked at the digital renaissance idea a lot. I am interested by the idea that the boundaries of organizations are starting to fall – the classical boundaries, like the classical constructive form. There is discussion about whether the traditional university will maintain its walls, and about open innovation and maybe the boundaries between the academic class and policy entrepreneurs. The public blogosphere is part of that phenomenon and there is a lot of hope and talk about a new intelligent software. There has been discussion about artificial intelligence forever, but maybe a more modest thing may come into place. You can see the start of it with digital smartphones. I could foresee these things becoming more powerful, the volume of communication getting higher, and these sorts of things routinely managing traffic, scanning the web for you, serving the secretarial function, hunting out huge volumes of information. The consequence would be people expecting to have this kind of limited intelligent agents working on their behalf.

How do you think this topic should be addressed by future research or policy? What fields and how should they approach this most successfully?

Meaning the fundamental discoveries that could enable this? I am interested in organizations as technologies. People are actively attempting to reengineer organizations to create high performance organizations, that are able to deal with uncertainty, transcend national boundaries, recruit talent to utilize smaller units of activity. So organizational scientists, gain theorists, market designers all these people are enabling a lot. I observe dramatic changes happening in machine learning. We could see artificial intelligence techniques, and although data mining techniques were oversold, there has been real progress there. I could easily foresee that coming forward. The real progress .has been in dealing with uncertainty: route functions that can be taught to computers to routinely deal with fuzzy, ambiguous, uncertain information. Internet sociologists can observe what has gone wrong with bloggers, and media scientists can study what has gone wrong with the evolution of the media and the internet. Education professionals are talking about the role of the university. Journals are attempting to open up the review process by posting things on the internet, and trying to find some balance of how to have informed commentary when making an open and transparent system. On the other side, the digital renaissance requires a certain level of material culture. You can see at least in the Netherlandsa deep concern with sustainable energy level systems and a European-wide electricity grid and markets. I think you could see signals of ever more electricity being traded across national boundaries. A natural consequence is mutual interdependency on electricity if that trend keeps coming. Likewise, there is stuff going on with manufacturing and I suspect that manufacturing will go more local, more low cost, more corner shop. One example is 3D printing, which is one of these technologies that are taking decentralized manufacturing to an unknown and unprecedented scale. 

Does this tie into the second wild card you were talking about: the ‘material plenty’?

Yes, because it increases inefficiency in manufacturing and there is discussion about it almost being a disaster. We have had huge advances in material plenty in the Western world, but an increase in unemployment in the manufacturing centre. There is a commentary about where those jobs are going. What happened to those jobs? Have they been taken away or was the loss of these jobs a result of the natural increase in manufacturing productivity and were they going to disappear anyway? So for Europe to achieve full employment, those people who were doing these jobs would be freed up from more routine tasks. The manufacturing centre has already been eroded, but carrying that still further, the question will be: can Europe utilize this talent that was working just to maintain basic infrastructure? If that becomes routine through infrastructure operating systems and a more digital society, what will people be doing? A positive use of those people would be engagement in creative and intellectual scientific activities.

What are the weak signals that could hint at the likelihood or imminent realization of these wild cards happening?

You could watch the natural miniaturization of electronic devices and track that to see how far that goes and how smart these things become. There is a lot of talk about web standards and the semantic web. The difficulty tracking that one is that is a diffused topic. Half the problem is creating a common standard. So even as the technology progresses, that does not necessarily mean that the standards progress. On the material plenty side, you could be looking at energy trading, electricity trading. You could watch the fraction of renewable energy used in Europe, the transnational trading of these things and the consolidation of regional markets to create a more widespread trade of electricity, thereby enabling more fluctuating sources of electricity, thereby stimulating more green energy. Tracking new institutional forms would be difficult, but maybe tracking some of these new things. We are all a little disillusioned with these new risk-bearing instruments that are blamed for the current downturn in the economy, but I think we have not seen the end of those. We will see a second wave of market place innovations and other sorts of firm-based innovations to enable or help us achieve more things with our firms. Maybe it would be possible to track mentions of these things in the popular news. Because of the sectoral basis of Europe, tracking employment by sector would point pretty clearly to these kinds of seed changes happening.

Would you like to add anything on the links between the wild cards and weak signals?

I conceptualize that and we did talk about it as new institutional forms and a rise in higher values. If we talk about a hierarchy of values, having the material values satisfied may see an increase in environmental values and esthetic values and curiosity as a value. Also the creating of a cognitive surplus of people displaced in a positive way from occupations where their range of ability was not fully being used. We will see more people doing a form of volunteering, e.g. amateur musicians who don’t need a recording contract to puttheir work on the internet. That sort of enabling of people without big labels to do what they like to do and share with others.

Looking ahead to the future of European research, which of the wild cards and weak signals do you think should be given top priority in European research?

It seems the ERA experts are primarily interested in research policy. so I would prioritise the digital renaissance. Because of peer review, maybe grant procedures will be changing, opening up grants from the monolithic research coalitions. The desire in policy is to create networks of researchers. The institutional capability of engineering networks, rather than keeping fixed constellations of actors, is not there yet, but I think it will change. 

We define a wild card as a low likelihood but high impact surprising event.  Would you agree with this definition?

Yes, your definitions were very interesting to me and my colleagues. We are engaged in a programme of research we call a deep uncertainty. We are talking about engineering systems and policy, and we are having a cross-Atlantic dialogue with Carnegie Mellon, who are very oriented to probabilities, fixed definable units of uncertainty. Many of my colleagues are talking about the big uncertainties that can’t be measured, where probabilities might mislead you. So we are influenced by Taleb’s Black Swans and we are talking about scenario analysis. We are also very interested in these wild cards, and the question is: how close are these definitions? The differences might surround whether or not we are more interested in the events surrounding the world or the impacts on the policy maker. In the end we need both. How is the decision maker affected by the uncertainty? What is the uncertainty itself? Although we don’t call it a wild card, I understand we are coming from a broad agreement on the value of looking at these things. As far as weak signals go, that is also a very interesting dialogue. We are talking about adaptive policy making, the kinds of things that you can monitor in the environment. A signal changes decisions, or shows that things are headed in the wrong way. So we categorise the problem. In the past, decision making was fixed and rigid: adopt this framework and follow it until it is time for a new decision. We need a more intuitive cycle and a chance to recognize and realign decision making, as the problems of the previous one become apparent, as certainties become resolved. We should help the decision maker to take advantage of these options, so althought it is stylized differently to weak signals, we definitely recognise that in our research. 

So you would broadly agree with both these definitions?

Even more than broadly. I am curious about how many other dialogues are out there using similar words and similar concepts. We touched on Black Swans and wild cards and adaptive policy making and sign posts. I have no doubt about the interest of those things, more a curiosity about whether they can be brought into a common framework to deal with uncertainty in decision making. 

Have you done any previous foresight studies using this approach with the wild cards and weak signals?

I personally have not. My colleagues dealing with flooding policy and water are using these kinds of things. Many of them are more infrastructure-related than knowledge- and innovation-related. Some of the interesting discussions are more to do with the ability of decision makers, and how can they manage the political realities of admitting we are in a certain environment. Similarly, engineering design has a tough challenge of making design trade-offs. They can recognize the uncertainty, but in the end they are going to need to make these trade-offs and how do we give information to developers and designers so they can incorporate this in their work? This is one of the reasons why they are very fixed with probabilities, which are important for some expressions of uncertainty but not others. They are driving towards probabilities all the way through, because it can be used in the design methodology. We also have colleagues working with RAND.1 They are doing what they call exploratory modeling. They use a simulation approach to understanding the future, but they acknowledge the uncertainties in the model. They try to do more model runs across a wider range of parameters, but they have the issue of how do you expresses the uncertainty and how do you act upon it? They hope by doing extensive enough simulations they can get round it, but you still have to come up with a metric. They are using things like regret from decision theory to say, ‘Faced with this uncertainty, how do we choose the best policy?’ It is not the optimum policy, but the one across all futures that provides us with the least regret. I find these discussions interesting – how to operationalize and act on uncertainty. Having argued so strongly against using probability approaches, they still have to synthesize all the data they are creating. That means they have to evaluate how to help decision makers chose the right futures. If you can’t just average them through created expected utility, what do you do about it? Taleb, in his book about Black Swans, was coming from a financial trading domain and seems to have no problem talking about expected utilities. He agrees that probabilities ought to be useful, averaging through is the right thing, deriving expected values, but he concedes that you will never collect enough data to do it. RAND, on the other hand, fundamentally disagrees that expected outcomes have any usefulness at all for the deeper uncertainties that we face.

What are the best methods to identify weak signals and wild cards?

Creativity techniques are the best way, particularly for your problem. What I like about the research problem is you frame it and you are searching the boundaries and limits of what we know. Therefore techniques that enable the opening of the mind, imaging a wide range of possibilities, are best, and other techniques that involve scenario analysis and exploratory analysis. Exploratory analysis has some merits for examining weak signals, because you can discover that even small changes in the signal could result in dramatic changes in outcome. You are freed up to explore wide ranges of values for the signal to reflect our current epistemic uncertainty about what the world is today, so that is also useful. It is appreciating that we can’t have a metric for a certain class of uncertainty. I appreciate the necessity for scenarios, which is a natural way of loosening up a parametric framework and attempting to measure and manage uncertainty in a less restrictive way.

When you say ‘creativity techniques’, do you mean group workshops?

Yes, and list association and free association techniques and brainstorming. You could make a distinction between the degrees of group work. You could also examine the expert vs. novice dimension there. Do you need an expert or someone who can be imaginative with the ideas, and how much of a consensus do you need and should you allow to form, or will it stifle ideas? The overlap with ‘expert opinion’ and group negotiation techniques – they are not the same, but they could enrich each other. If I were to design a group negotiation or group learning exercise here, I would attempt to find the widest variety of people possible. Certainly some of the decision makers, but also some people who are just attempting to think creatively but publicly about the future. In particular, if we were going to talk about the digital renaissance, it would be nice to have people who are talking about the future of the internet and the bloggers and the people who are shaping opinion and talking about these things online in the digital world. That also raises the issue of whether the exercise should be digital, because maybe the people who who are shaping opinion and talking about these things online in the digital world. That also raises the issue of whether the exercise should be digital, because maybe the people who care most about these things are online discussing them.