A few years ago when we were advising the government on opening up data, we knew that we wanted an organization not of the government but one that could inform the government. An organization that could promote the creation of a trusted data ecosystem for everyone, while acting independently and highlighting the societal and economic value from open data. This led to the creation of the Open Data Institute.
COVID-19 has clearly demonstrated that in the absence of a healthy and mobile population, the economy starts to fall apart. Ensuring that public and private interests are mutually supportive has become paramount today. If we look at the UK’s Office of National Statistics (ONS), the data they collect and publish has become essential to our efforts to cope with the pandemic. It includes data on the number of deaths, where people live, their economic circumstances, where they travel, etc. All of this holds immense potential for understanding the pandemic and developing responses to it. So, it is extremely important to democratize data and make it available in the way that organizations like ONS do. Governments have a fundamental role in setting up institutions where information is made available as a public good.
Meanwhile, the private sector also has a responsibility. It’s a public-private relationship at the end of the day. Thus, the private sector needs to be advocating for and promoting some of the standards that will help bring about interoperability. It must be involved in the processes to ensure that the public infrastructure works, and can then connect that with its own infrastructure. We already see this kind of symbiotic relationship in areas like healthcare and logistics. Making data open and treating open data as infrastructure is a great way to try and tackle a wide range of concerns form the digital divide to promoting innovation.
Building sustainable data infrastructure
Committing to open data infrastructures also prevents the formation of any kind of a monopoly that ties people into proprietary formats or a particular kind of standard. It’s especially beneficial for developing countries, as often they aren’t hindered by legacy proprietary systems and can therefore move quickly. The cost of maintaining accurate geospatial data has also gone down significantly. This presents an excellent opportunity for developing countries to bring in new and disruptive ways of collecting data, while assessing how it is applied across the range of challenges they face. This will also allow them to be more innovative. An open data manifesto is a great way to ensure that developing countries can construct a more cost-effective and sustainable data infrastructure in the future.
With the evolution of Artificial Intelligence (AI) and advanced computing, algorithms are becoming more and more complex, and understanding the underlined principles at work that give rise to the outputs is much more difficult. Earlier, decision support systems were programmed in a rule-based way, and the rules that were applied would together offer an accessible form of explanation. But for modern large-scale neural networks, explanatory adequacy is a challenge. Trust is a key factor for the effective deployment of AI — today developing explainable AI is a key area of research.