The Edinburgh Conversations saw military, political and academic figures from the Soviet Union, United States and Britain discuss big challenges, seek to ease tensions and reduce the prospect of a nuclear catastrophe.
Fast forward to 2019 and Edinburgh is positioning itself at the heart of a new series of critical global conversations – this time around data and how it is used, as we grapple with big ethical questions in a rapidly-changing world.
The Centre for Data Ethics is taking shape in the Edinburgh Futures Institute (EFI) as it prepares to emerge from inside the landmark Royal Infirmary building, opposite George Heriot’s School at Lauriston.
Baillie Gifford, the global asset management business, has supported the development with £5 million. This will include the creation of the Baillie Gifford Chair in Data Ethics and a research programme into ethical issues around data and artificial intelligence (AI).
When it opens in 2021, the EFI will be one of the largest centres for interdisciplinary learning and research in Europe, with a key responsibility to engage the public in the challenges and opportunities posed by data and AI. It is just one aspect of a push to make Edinburgh the data capital of Europe, following the creation of The Data Lab innovation centre in 2015 and now the Data-Driven Innovation (DDI) initiative, part of the Edinburgh City Region Deal.
The DDI initiative is supported by more than £660m of funding from the UK and Scottish governments and academic institutions (Heriot-Watt University and the University of Edinburgh) and has five ‘data hubs’ – including the EFI – designed to drive economic and social benefits from the data revolution.
Lesley McAra, director of the EFI and assistant principal of community relations at the University of Edinburgh, explains: “EFI starts by looking at specific challenges, not specific disciplines. We are building teams that have never existed before.”
McAra says the new centre is crucial to the future of data in Edinburgh: “Ethics is at the heart of what we are doing. We have to get that right.”
The Baillie Gifford money is funding 18 PhD scholarships, which the university is matching (for a total of 36).
“There will also be two Chancellor’s Fellows, who are stellar researchers coming in to drive things,” says McAra. “The Baillie Gifford money is a wonderful gift which has catalysed things and allowed us to develop the project much more quickly.”
The chair of the new Centre for Data Ethics is due to be appointed within the next few weeks, in post from early next year and inside the new EFI building when it opens in 2021.
“The centre is about research and development, education and regulation and how we manage data,” says McAra, pictured. “It’s about primary research and working with external companies to ensure they are looking at data ethics, but it’s also important to the university itself and how we operate.
“There is also outreach in terms of public trust and confidence. Building public engagement and confidence – and understanding – in the way we all handle data is crucial.”
McAra says the centre has global goals too: “I want Edinburgh to be seen as the place to go to tackle big ethical challenges around data. We could draw on the history of the Edinburgh Conversations and be the place where groups can convene to discuss really big issues involving data such as the impact of climate change, the regulation of tech conglomerates, the rise of mass surveillance and loss of public trust and confidence.
“We will not be afraid to take on big challenges. Edinburgh and Scotland has the power of convocation; the history of the university with the Edinburgh Conversations and the progressive nature of Scotland means we have the power to be that convener. The university has a big leadership role to play.”
Gillian Docherty, executive of The Data Lab, agrees: “Part of that leadership role is collaboration between the public and private sector and academia to help maximise the data opportunities for Scotland. I think we are doing pretty well; on our travels across the UK and globally, we get feedback that the public/private/academic trinity is very difficult to crack, but people see we are doing a pretty good job here.
“We have lots of examples where collaboration has helped us to leverage data and derive useful insights – and where we have done that in the right way.”
Jarmo Eskelinen, director of the Data-Driven Innovation (DDI) initiative, says collaboration is a must to succeed in data: “Edinburgh and Scotland are part of a wider data ecosystem. A sizeable chunk of the innovation in data science, computer science and informatics has come from the UK. Edinburgh is in a good position and we need to take further advantage by working to maintain a balance of co-operation and competition with other European centres going forward. We have to be part of that network and we can certainly do bilateral deals with other centres. We are going to Helsinki in September to sign such a deal and others will follow.”
It is often said that Scotland is a village, which aids deep collaboration, and Eskelinen agrees: “Edinburgh is just the right size of city to sit in this data space. When it comes to innovating in tech, it’s pretty unique. It’s a small city with a very good quality of life which attracts people – from its own universities and further afield – and companies. It has a strong, highly skilled workforce which provides a talent base of a much bigger city. There is a legacy of data innovation and service innovation in Edinburgh. The Centre for Data Ethics is the latest component of a very dynamic period of growth for the city – and part of a wide-ranging response to these big challenges around data.”
Eskelinen says it is crucial the new centre and university join forces with the city, city region and Scotland as a whole: “We need to work with the public and private sector to create a test-bed to develop exciting new products and services and use data science to address in a collaborative way the big problems and challenges that face us as a society. It’s about identifying challenges and bringing a wide range of people together to try to tackle them.”
He also thinks there must be deep collaboration across different data disciplines: “Data science is a hip profession, but it’s just one link in the ecosystem of understanding data. Before data science, you need data engineering and that needs to be just as conscious of the issues as data science. We need to ask ‘What kind of data is this? Is it good quality? Does it address all aspects of this particular topic we are looking at?’ And as well as data engineers, there is a crucial role for statisticians to help us understand the bigger picture from a wide range of data sets and make informed decisions.”
As Eskelinen says in his first person comment in this supplement, balancing the triangle of innovation, regulation and privacy is tough – because regulatory and legislative processes are slow to react to the pace of innovative change.
But he believes the new centre and its research power can help get the balance right, by providing the insights needed to develop policy.
And while the desire to make Edinburgh the data capital of Europe has been voiced widely, there is perhaps an even higher prize at stake – making the city the data ethics capital of the world.
GETTING REAL ABOUT ARTIFICIAL INTELLIGENCE
If the pace of change in the data world is tough to comprehend, the dizzying speed in the growth of artificial intelligence (AI) is an even greater challenge.
Jarmo Eskelinen, director of the Data-Driven Innovation (DDI) initiative, describes it as a “fast-moving target” and has concerns about whether regulation and legislation can ever keep up.
“The problem is that traditional frameworks are regulated but AI frameworks are not – and as AI starts to think for itself, we do not know what happens,” he says. “As AI develops, there are many unknown unknowns.”
Firas Khnaisser, chair of Data and Marketing Association Scotland, says there are real implications for people’s lives: “We’re heading into a future powered by AI, where our lives will be run by algorithms. When things go right, they will go wonderfully right and when things go wrong, they will look like something out of a dystopian sc-fi film.
“What that means in real terms is that if we are incorrectly profiled by an algorithm, we could lose access to products, services and at times even our freedom.”
When it comes to biases in AI, Gillian Docherty, chief executive of The Data Lab, says: “We are training algorithms on biased data so therefore the algorithms will be inherently biased.
“I think the whole space of open and explainable AI will become much more important.
“We are not in a position where we just accept ‘the answer is the answer’ from AI. If the computer says no, we have to be confident the decision has been made based on unbiased data. There will be a lot of research and a lot of economic opportunities in this space.”
The Scottish Government recently tasked The Data Lab to convene a process to develop an AI strategy for Scotland.
“We are at a very early stage,” says Docherty, “but we need to ensure it is inclusive and citizen-led.”
EXCEPTIONAL COMPUTING POWER
The Data-Driven Innovation (DDI) initiative is a ten-year programme which aims to help all citizens and organisations in the Edinburgh city region benefit from the data revolution. The DDI initiative is supported by £661 million of funding from the UK and Scottish governments plus Heriot-Watt University and the University of Edinburgh and has five ‘data hubs’ – including the Edinburgh Futures Institute.
Its website describes its mission like this: “Through high-speed data analytics, our ability to capture flows of data and understand what they tell us is bringing better and faster capability to identify trends and behaviour across many sectors, leading to improved services for consumers and citizens.”
Experts from the universities of Edinburgh and Heriot-Watt will collaborate with industrial partners on data-based projects in the public, private and third sectors throughout the programme. It aims to “increase the contribution of university research and in-demand graduate skills to the region’s economy, launching more spinout companies, attracting start-ups and established businesses, and driving public and private sector investment”.
Five DDI ‘hubs’ have been created to help ten industrial sectors become more innovative through data.
The University of Edinburgh hosts the Bayes Centre at its Easter Bush campus in Midlothian and the Usher Institute – as well as the Edinburgh Futures Institute. The National Robotarium, pictured above, is a collaboration between Heriot-Watt University – where it will be sited – and the University of Edinburgh.
Supporting the work of the hubs is the new £79m supercomputer Archer 2, which will replace the earlier £43m Archer system, unveiled in 2013 and capable of more than one million billion calculations a second. Archer 2 is designed to deliver “secure and trustworthy analysis” of huge amounts of data on a scale which will be unique within Europe.