Expectations and realities – part 2

In part 1 we thought about what we expect from innovation on one side, and how it evolves in reality on the other. It became pretty obvious that there is a significant mismatch between the facts about innovation and what we think them to be. In this second part now I’ll try to cast some light on the underlying reasons for that discrepancy.

As Amara’s Law famously describes it, “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” There seem to be two separate forces at work, depending in the time frame we consider. In addition, and as described in part 1, Brian Arthur in his work on The Nature of Technology clearly identified major differences between individual technologies and technology domains. Individual technologies are developed for a clear purpose, usually through standard engineering procedures, along a timeline that is measured in months to a year. In contrast, technology domains essentially emerge without a definite purpose; they co-evolve together with society and its institutions through mutual adaptation; this complex process typically measures in (several) decades. With this distinction in technology development processes, it makes perfect sense to address the two different timescales separately.

First – the short-term, the domain of individual technologies. According to Amara, we are too optimistic about their impact. Why could that be? I’d offer two arguments. One is a simple question of time. Individual technologies are mainly generated through standard engineering practice, so that should be pretty predictable, right? Well, not quite. The initial idea might be straightforward, but remember that a technology is the combination of components, each of them itself a technology. So implementing a new idea means to make sure that all the interactions between all the components work seamlessly. In this complex setting, it’s the proverbial devil in the detail that mocks our expectations.

My second point regarding the short-term is about optimism, or positive bias. We tend to focus on our positive intentions, hence ignore the collateral effects our innovation can have on others. Because individual technologies are usually fairly concrete and tangible, and because they are the result of purposeful effort, we tend to believe that we can impose our intentions on the outcome of our effort. We seem to be blind for the fact that unintended consequences do exist. Therefore, we fail to account for them in our expectations, which are unrealistically positive.

Second – the long-term, the domain of technology domains. Here, we are too pessimistic about their impact. Again, why is that the case? Brian Arthur reminds us that we are talking about a fundamentally different concept of technology when we tackle the long-term. Technology domains (or groups, families, clusters of technology, such as biotechnology of information technology) co-evolve together with society through a complex process of mutual adaptation. I’d argue that (a) there is not a lot of awareness for this specific process; and (b) even if this was to become common understanding, the process itself, through its complexity and long-term horizon, is too big for us to grasp, is beyond our imagination. I feel that we don’t have the intellectual means to fully understand and ultimately predict that process and its outcome realistically. A pretty sobering thought, I must admit.

Regardless of such a moment of humbleness, we still have expectations for the long-term, don’t we? How do we shape them? Or rather, how do they take shape? My theory goes like this: because we don’t have a view of the complex co-evolution of a technology domain and society (see (a) above), we actually take only a part of the domain into account: the individual technologies that are visible early on when the domain begins to take shape. That implies that we only see the immediate purposes of those single technologies, and that we judge the impact of the domain only as the sum impact of the known components of the domain. That view misses two important elements of what defines the full potential of a technology domain. Because of the fairly slow progress of the domain, we can only see the “early” components of the domain. As it evolves and matures, it will entail more and more components that we can not imagine today. Secondly, the co-evolution together with society implies that the domain finds additional purposes as society and its institutions adjust to the emerging domain. This might seem academic, but it is essential for how limited our expectation of the long-term effects actually are: we miss a full view of the single technologies that will build the domain, and we miss a full view of the purposes these technologies can fulfil. But it’s the linkages between technology and purpose, between purpose and technology, that define the full potential of a technology domain. Hence our limited view of what a domain is and how it works makes us blind for the substantial impact technology domains have in the long run.

This might come across as a saddening account of our collective failure to appreciate the impact of innovation. But what can we do about all that? I see three interconnected themes. The first is simply awareness of our biases, so that we can take them into account when we plan innovation activities. The second is an element of general education towards what I call innovation literacy, the understanding for what innovation is, how it works, and why it is relevant. Thirdly, I see a need for innovation policies that account for the biases, and that promote the education part. That will require long-term engagement. And I guess that I underestimate its positive impact …