Inside the Information Society returns after a break. This week’s post (based on a speech written for a recent conference) looks at the relationship between innovation and development.
Complexity lies at the heart of the challenges that face those of us who are concerned with innovation and development. We need to pay more attention to that complexity rather than embedding it with false assumptions.
Challenging the Whig interpretation of technology
The relationship between innovation and development is never simple. One of the most important observations in the World Bank’s review of Digital Dividends two years ago was that ICTs have not delivered developmental gains on the scale that had been hoped. One reason for that is that too many people have assumed the link between innovation and developmental outcomes is automatic.
Historians have a term for the idea that human experience is an inevitable progress from bad to good, poor to rich, dictatorship to democracy. They call this ‘the Whig interpretation of history.’ We who are concerned with innovation and the Internet should avoid the Whig interpretation of technology.
There is no automatic link. Innovation interacts with highly complex economies, societies and cultures in ways that are anything but simple, and it takes much more than hope and infrastructure to achieve developmental goals. It takes serious understanding, public policy and cooperation between different people and different institutions.
I will make five broad points about this from a public policy perspective.
There are no simple solutions to developmental problems
First, innovation must be understood within its context, in particular the underlying structural challenges that have inhibited development. I’d argue strongly here against the word “solutions” which is so beloved of many advocates of ICTs.
Information technology doesn’t provide solutions to deep-seated structural problems of poverty, inequality or conflict, but it can make a contribution to mitigating or reducing them.
Its contribution isn’t necessarily benign. Innovation may be broadly beneficial overall, but there’s nothing inherently beneficial about individual innovations. Some of the most successful innovators have been criminals; some work for dictatorships. We need to think carefully and act accordingly if we want to emphasise the positive.
Nor is every innovation an improvement. Policy pronouncements often call for “innovative solutions” to problems for which tried and tested models – or indeed yesterday's innovations – already offer "remedies". The new can be the enemy of the good.
As the World Bank pointed out two years ago – and as now seems increasingly to be accepted – ICTs so far have probably increased inequality rather than reducing it, because they’re more available and can be more effectively used by those that already have income and capabilities than those that don’t.
In any case, most impacts of most new technologies are unexpected.
Don’t confuse ‘change’ with ‘development’
ICTs and innovation don’t lead to development alone, in other words; they lead to change. It’s the interface they have with public policy that matters if we want to make that change development.
So the central question here is this: Do we let innovation shape our future, or can we act to shape that future, making use of innovations (plural) in ways that we think best? If it’s the latter, then:
We need to think through the implications of changes that are taking place today and coming next.
We need to understand and anticipate their impacts on economy and society.
We need to avoid assumptions that all will be well, and assess the risks at least as much as we assess the opportunities.
We need to develop policy approaches to innovation that are likelier to achieve the outcomes we desire.
What’s the goal?
So, second, how do we define these outcomes? One way we might begin is to recall the opening sentence of WSIS’ Geneva Declaration – for a ‘people-centred, inclusive and development-oriented Information Society.’
Not before time, I hear some echoes of that in language that’s now coming from within the ICT and innovation sector: language concerning ‘human-centred’ technology and innovation, for example ‘human-centred artificial intelligence.’ There’s recognition in this of growing disquiet among citizens (and governments) of the risks in letting technology take the lead in determining the future.
The obvious context for international organisations to use here is the sustainable development goals which emerged – it should be emphasised – from a much earlier set of internationally agreed objectives than WSIS: objectives that are based on the importance of assuring three things – economic prosperity, social welfare or social equity, and environmental sustainability.
The fact that sustainable development’s an older goal than those agreed at WSIS doesn’t make it less important but more fundamental. We should be looking for ICTs to serve SD, not vice versa as seems to be the case in some of the more enthusiastic advocacy literature.
Innovation and sustainability
This integration between ICTs and sustainable development requires more thought and more interaction between their separate paradigms. In particular, it requires us to think afresh about the interface between their regulatory approaches – the ‘permissionless innovation’ that’s beloved within the Internet community, and the ‘precautionary principle’ which underpins sustainable development and is the norm in almost every other sector.
(An interesting example here is cryptocurrency. Bitcoin may be awfully clever innovation, but it’s energy demands are not sustainable at scale. Innovation practice that paid attention to environmental impacts would have pointed its developers in more sustainable directions.)
Widening discussion about the implications of technology
Third, and therefore, more dialogue is needed between those who are concerned to innovate and those who are concerned with the economic, social, cultural and political impacts that arise from innovation.
A lot of discussion about ICTs in development has been too narrow. It’s focused on what individual programmes, projects or applications might achieve. What’s more important, often (probably most often), are the systemic ways in which ICTs have altered and will alter power structures, business dynamics and citizen behaviour. These are often very different from what technologists and policymakers expect or hope.
Systemic changes are taking place in most places as a result of how innovation(s) is/are becoming used. All areas of public life are affected in ways that weren’t anticipated, some positive, others requiring “remedies”; and they'll be more affected in the future. Our legal and regulatory frameworks are too slow to address these changes because innovation and adoption of technology are too rapid for the way they work. Not just rapid now, but accelerating, which makes it increasingly difficult to shape them towards ‘people-centred’ outcomes.
Where decisions are being made
The locus of innovation power is also changing. Once it was geeks in labs that made decisions that designed the Internet. Now it’s the boards of global corporations – or in some cases individual business leaders. Technocratic power is concentrated both commercially and geographically. The Internet and ICTs have become much more like other major economic sector (oil, mining, pharmaceuticals, vehicle manufacturing) than many of their advocates like to admit, and problems arise as a result (for example those where privacy conflicts with commercial interests).
What’s missing from the “multistakeholder” environment?
Fourth, what then of multistakeholder participation? That’s been an important theme in governance of the Internet, one that was further stimulated by WSIS and that’s become a norm in much discussion around the governance of innovation generally.
It’s important to remember that the case for multistakeholder participation is about the quality of decision-making that it supports. Ensuring that a wide range of expertise is drawn into the process, and that those affected by decisions have their voices heard: these two (notably different) types of inclusion, it’s anticipated, should lead to decisions that are better.
But there are risks in this as well, as it plays into power structures old and new. One is that it can foster decision-making by insiders, what I’ve called elsewhere ‘multistakeholderism of the like-minded’, making decision-making less not more inclusive.
It should be about dialogue between those with different interests and perspectives, that’s aimed at shaping the future in the interests of the whole community, rather than just another way of contesting for decision-making power now. Its spirit, logically, is compromise not contest.
Multistakeholder engagement adds value but two things are needed for it to do so.
And that dialogue should be inclusive, not just a dialogue among those in the know. As Fei-Fei Li, chief AI scientist at Google, said in a recent Oxford lecture, innovation won’t be successful if it only involves computer scientists. They need to be in dialogue with anthropologists, economists and social scientists – experts in fields that deal with humans rather than machines. I’d make three further points about inclusion.
It must include women equally with men, and indeed all under-represented groups within society. Innovation that’s undertaken by young men in California will be innovation that suits young men in California (more than or rather than the rest of us).
It must include end-users. I heard an academic in this field say recently that the views of farm workers were less important here than those of knowledge workers. No, they’re not. Development and innovation are sustainable only if their beneficiaries (or victims) are involved in the design of innovations and the policies around them.
And it must include the global South far more than is the case at present. It’s fifteen years since I and colleagues wrote the Louder Voices report about under-involvement by the global South in ICT decision-making. Not enough has changed.
What about regulation?
My fifth and final point’s concerned with regulation. There are many regulatory challenges, not least that (mentioned above) between permissionless innovation and the precautionary principle. I’ll point to three further themes that need to be addressed.
Regulatory challenges are complex and multifaceted. Among those that I cited in a recent blog are technical standards, competition policy, data protection, content, cultural autonomy, the impact of innovation/ICTs on politics and power structures, and taxation. These are all interconnected. What’s done with one affects them all.
Circumstances are changing rapidly, and ever more so. The pace of innovation is faster than that of policy development or law or regulation. Are there ways in which regulations can be made adaptive, shifting their nature as technology and markets change, rather than depending on decisions that can only ever be behind the game?
We need regulations that will work with hybrid contexts, especially in the face of artificial intelligence. We have rules for contexts we can call ‘before AI’. We can imagine others for AI-enabled worlds. But it’s the rules that we develop for hybrid contexts in between that will influence technology’s development and future outcomes.
Potential, caution, context
When I first wrote about ICTs and development thirty years ago, I emphasised three things: potential, caution, context. Those remain key themes for innovation policy today. Issues at the interface of innovation and public policy are challenging. They need to be addressed with due attention to opportunity ('potential'), risk ('caution') and diversity ('context') if we are to move towards a ‘people-centred, inclusive and development-oriented Information Society.’
Image credit: Innovation, on Unsplash.com