But the power of data also drives political populism and extremist movements – and contributes to climate change due to exponentially rising power demands of always-on, cloud-based services.
“Doing data right” means to understand different aspects of the data and systematically reach for balance between them. It is no use just wringing our hands and asking if we can ever really keep up with developments in data science and artificial intelligence (AI).
One of the biggest challenges in doing data right is the challenge of speed. Innovation in data processing and artificial intelligence is so rapid that control systems are often left far behind. We need to focus on agile regulation, ownership of data and data ethics.
“Agile regulation” is an oxymoron – regulation is, by default, rigid and slow to react to the pace of innovative change. Still, we must learn to regulate faster. One way to achieve this is to use experiments like companies which develop services by prototypes and fast iterations.
Policy-making requires decisions based on proper analysis. Real-world experiments in living labs can provide data which helps derive those crucial insights to decide how to develop policy.
Data holds immense value, but whose data is it? “Sharing data increases the value of data” is a cliché often repeated. However, instead of sharing, corporations are monopolising data, like any critical asset. How can we ensure that data doesn’t just congregate around a small number of the largest companies and governments but delivers public good? How do we protect the data rights of individuals?
Legislation alone is too slow and too blunt a tool. Data commons, data trusts and easy-to-use tools to manage our identity are ways for us to see and control the use of our data easily. Self-sovereign, transportable identity would let the users be in control of their data across all services they use.
The public sector should also be aware of the value of the data it creates and holds. Innovations in health need health data, climate solutions environmental data. We need business models which deliver value for all stakeholders, enabling innovation in public services.
Lastly, the challenge of speed is especially acute in data ethics. AI and machine learning are increasingly making decisions for us. Classical ethical frameworks are of limited use when we do not know how machine learning systems process data and make those decisions.
The world is determined to close the gap, as the numerous new initiatives in data and AI ethics prove. We need to balance the triangle of innovation, regulation and privacy. How do we create data ecosystems which are ethically sound and provide opportunities for both companies and individuals?
Doing data right is one of the main goals for the University of Edinburgh in the Data-Driven Innovation initiative. We are creating the Centre for Data Ethics as part of a dynamic and wide-ranging response to these significant challenges.
We are in an exciting place in Edinburgh. We have the chance to be right at the heart of understanding how to do data right. The city is the leading UK and European research base in informatics. As part of the European Union, the UK has been a member of the only regulatory entity which has repeatedly taken action against monopolistic data companies.
The other global models out there are very different – the government-controlled model of China, or the US oligarchy, driven by huge digital companies. Europe hasn’t always got it right, but it’s a legacy to build on because it has always respected the rights of individuals and smaller companies.
Even if that pace of change is quicker than it has ever been, we have faced similar challenges in responding to technological progress since the Industrial Revolution began. Rapid change is a never-ending process, and the best we can do is be prepared and reactive.
I’m optimistic about the future. The power of data can help us to crack some of the biggest challenges we face – in terms of sustainability, inequality and even the current upheavals sweeping across Western politics.
The pessimistic view is that we fail to rise to those challenges.
Data plays a significant role in both of those future scenarios. If we do data right, it is much more likely that the optimistic scenario will prevail.
Jarmo Eskelinen is director of the University of Edinburgh’s Data-Driven Innovation initiative