How do we keep data and digital innovation ethical?

David Lee hears about how the University of Edinburgh is helping connect academia, industry and society to use data ethically and effectively

Working in collaboration. The University of Edinburgh’s Data Driven Innovation initiative is welcoming all manner of input to further its goals. Picture: Shutterstock
Working in collaboration. The University of Edinburgh’s Data Driven Innovation initiative is welcoming all manner of input to further its goals. Picture: Shutterstock

How do we want to use data? Do we want to utilise it to make financial services more open and accessible to all? Can we harness it to examine the flaws of our elderly care system, laid bare by ­Covid-19, and try to make fundamental improvements?

And do we want to ensure that the march of data and artificial intelligence in shaping all aspects of our lives is managed in an ethical way?In the last year, the University of Edinburgh’s Data-Driven Innovation (DDI) programme has answered all these questions with an emphatic yes, furthering the Edinburgh city region’s ambition not to be just the Data Capital of Europe, but a world-leading hub.

Professor Jonathan Seckl, vice-principal of the university, is one of the academics behind the 2016 Science and Innovation Audit, which ultimately brought the DDI programme to life.

"We are delivering transformational research, innovation and support to society across the board and to the highest possible ethical standards"

He says: “We saw in 2016 that data and digital innovation was strong, but thought it could be transformational. We made a bold statement, and outlined the huge economic potential.

“It was a big step but we haven’t regretted it. You cannot understate the importance of the DDI initiative to the university and the wider region – and our clear focus is beginning to bear fruit as we reach into a series of critical sectors.”

Despite the dislocation of a global pandemic, the DDI programme has made very significant progress in the last 12 months. In all cases, partnership working has been crucial.

Last October, Professor Shannon Vallor arrived from Santa Clara University in Silicon Valley as the first Baillie Gifford Chair in the Ethics of Data and AI. She leads the Centre for Technomoral Futures, a ground-breaking collaboration between ethics and technology, which will make its permanent home in the under-construction Edinburgh Futures Institute (EFI).

Shannon Vallor

In January, a £20 million partnership to develop the Advanced Care Research Centre (ACRC) was announced between Legal & General and the university, with data fundamental to the first holistic UK research programme into elderly care. Three months later, the ACRC’s plans were accelerated as a result of Covid-19.

In June, another £20m-plus funding package was confirmed for the Global Open Finance Centre of Excellence (GOFCoE) in Edinburgh.

A further significant collaboration was announced in July, with Hewlett Packard Enterprise confirmed as the university’s partner in creating the Edinburgh International Data Facility.

The DDI programme’s response to the pandemic was also built on co-operation. A small grant fund of almost £250,000 was established for rapid-response projects which used data to tackle some of the huge range of challenges arising from the coronavirus pandemic.

These are covered in more detail on pages 9-11, but included collaborations with small fintechs such as Wiserfunding and CogniHealth and large tech companies, including Sopra Steria, as well as specific projects with partners like NHS Lothian, Scottish Government, The List magazine and social enterprise Breadshare.

Seckl is energised about all these projects, large and small, within an ever-widening Edinburgh ecosystem of data innovation.

“If used responsibly and carefully, data is transformational,” he says, “and trust is the first word in everything we try to do.

“The thought leadership we are showing with Shannon Vallor’s appointment as Chair of Data and AI Ethics, is absolutely crucial. Maintaining public trust [in data and AI] is vital.

“We are delivering transformational research, innovation and support to society across the board and to the highest possible ethical standards. Where better to do that than in the care sector, where this atrocious pandemic has swept in and revealed such long-standing problems and inconsistencies?

“Care really is the Cinderella sector and with the ACRC, we have an opportunity, working with a big company like Legal & General, to transform how we do it.”

The Global Open Finance Centre of Excellence (GOFCoE) is also founded on collaboration, taking Edinburgh’s open banking expertise into a broader sphere, with multiple partners involved, while further fusing the region’s twin excellence in both data and fintech.

GOFCoE is supported by a £22.5m grant from UK Research and Innovation’s Strength in Places Fund, and was devised and created by the university, alongside the Financial Data and Technology Association and FinTech Scotland.

“GOFCoE is complex, but fundamentally it’s about a new way of using financial technology to deliver financial services which are more accessible to everyone – and provide clear public and social good across the globe,” says Seckl.

“All these projects add up to a big statement from the university, and social good is absolutely vital to our mission. We are here primarily to educate, but also to research and innovate – and in doing that, to deliver fundamental human benefit. We are driven by a desire to improve our society in south-east Scotland and the world; that’s what motivates us all.”

This global goal underlines the crucial need for meaningful public-private partnerships, as well as the ability to attract international talent, like Prof Vallor.

So what was it about Edinburgh that convinced the eminent philosopher to leave California?

She says: “I have been championing for a long time efforts to find ways of fusing artificial intelligence, data and ethics, which is not just a mashing together of different approaches and methods, but something truly collaborative.

“Edinburgh Futures Institute was designed as an environment where new, multi-disciplinary methods of teaching and research were possible. From teaching spaces to the design of courses, to research groups, everything is designed with that co-operation in mind. What I had been saying needed to exist was being created and I wanted to be part of it.

“How can we move away from an antagonistic approach, to create something which isn’t a battle between technology and ethics, a clash of incompatible approaches and methods – but something truly collaborative?

“This is a real opportunity to fuse technology and moral intelligence from the ground up.”

Vallor adds that Scotland’s long and strong history in common-sense philosophy, as well as Edinburgh’s modern power in AI and informatics, was also a big draw, saying: “The opportunity to work in a university that had the tech capability was really important to me.”

Seckl admits such collaborative approaches were not always the university way – but he says now there is no going back.

He says: “Universities were a bit like Medieval monasteries, with professors and students in splendid isolation in their ivory towers.

“I exaggerate for effect, but there was a sense that universities were not part of the mainstream world. Now we work in partnership with businesses of all shapes and sizes – from Royal Bank of Scotland, Legal & General, Hewlett Packard and Baillie Gifford, to tiny start-ups – and with a whole range of public sector bodies and trade associations. This is not just about the university any more.

“In the Bayes Centre [the first DDI hub to open its doors], academics and public and private sector stakeholders work together in the same space. Students oscillate between science and the problems of industry.

“This is a new concept of what a university is and we are already beginning to see this holistic approach work. It’s a real change in university thinking.”In this context, what is the scale of the ambition for the University of Edinburgh and the DDI programme?

“Rather immodestly, we said we wanted to be the Data Capital of Europe.

“I’d like to go further; I want us to be one of the world’s leading centres,” says Prof Seckl.“We want to use data to totally rethink society – to help create a society that is fairer and more equitable, more balanced and more accessible – to deliver

better outcomes in a complex world. I see the DDI programme at the forefront of how we reset things.

“The world will not go back to how it was after the pandemic – and we want to be a powerhouse of the new thinking.”

Edinburgh International Data Facility panel

The desire to make Edinburgh a world-leading hub in all things data is underpinned by the creation of the Edinburgh International Data Facility (EIDF), with global tech giant Hewlett Packard Enterprise confirmed as the university’s partner in July.

The £100 million project will create “a place to store, find and work with data of all kinds – offering services for the long-term hosting and preservation of digital data and for state-of-the-art analytics and data science”.

The EIDF aims to drive greater collaboration between industry, the public sector and academia to deliver benefits.

EIDF is being created on the outskirts of the city by EPCC (formerly known as the Edinburgh Parallel Computing Centre). It will support the DDI’s innovation hubs – Bayes Centre, Edinburgh Futures Institute, Usher Institute, Easter Bush Campus, and National Robotarium – to enable research and development on global issues including food production, climate change, space exploration, and tailored healthcare.

Due to be operational this autumn, it will offer researchers access to high-performance computing and AI technologies, supporting them to apply analytics to modelling and simulation to increase the accuracy and speed of results.

Mark Parsons, director of the EIDF, said: “We believe EIDF is the only facility of its kind in Europe focused specifically on data-driven regional growth.

“Our goal is to collect and curate a large number of interesting datasets and make them ‘analytics ready’. Some of these datasets might be small; we hope many will be truly large … We’ll be working on making them as useful as possible – easily findable, accessible, linkable and interoperable.”

EIDF is operating a “building block” approach, with an initial ten petabytes of storage (1 PB = one million gigabytes) which will grow over time, driven by the needs of students, researchers and innovators.

Partners panel

John Godfrey, corporate affairs director for Legal & General and partner in the Advance Care Research Centre: “We know there isn’t a single answer to make care better – but we’re setting the bar very high and hoping for the emergence of whole new methods of care, based on sustainable healthcare and financial models.

“Data is fundamental; any changes must be strongly evidence-based. If we don’t know the full picture, at an individual or population level, we cannot decide how to improve things.”

Nick Thomas, partner at Baillie Gifford, which created the first Baillie Gifford Chair in the Ethics of Data and AI: “We are convinced that Edinburgh University has a world-class data science capability. We also admire its ability to blur the boundaries of academic disciplines, which will be necessary to make progress in the field of AI ethics.

“Interesting discoveries are often made when ideas from different fields collide. We aim to be a good partner for the Institute, making a long-term commitment to funding the Chair and embracing an open exchange of ideas.

“Some of the development and use of AI will be driven by companies in which we invest, so we hope to make constructive connections and work with Professor Vallor to include ethics at the centre of the discussion.”

Lee Rand, director of high-performance computing and AI at Hewlett Packard Enterprise, and partner in the Edinburgh International Data Facility: “HPE is proud to embark on this long-term initiative with the University of Edinburgh. In the data-centric era, deriving insights and value from across multiple datasets will be a key to success for business and government alike. We look forward to boosting the UK’s capacity for data-driven innovation through this initiative.”Gavin Littlejohn, global chairman of FDATA and partner in the Global Open Finance Centre of Excellence: “It is becoming increasingly clear that the grand open finance challenges GOFCoE will address simply can’t be solved by working alone.”

‘Technology and AI are human all the way down’

Professor Shannon Vallor explains why she wants to fuse the best of hi-tech advances and ethics to create “a future worth wanting."

We must avoid technological determinism – the idea that technology leads and society merely follows.

That’s a lie, and a convenient evasion of responsibility for those building their values into these technologies.

Technology and Artificial Intelligence are human all the way down – built to promote, optimise or systematise, to create power and realise specific values in the world.

Humans are the creators of this technology and we need to ensure that human accountability isn’t lost.

My work at the Centre for Technomoral Futures (part of Edinburgh Futures Institute) starts with the premise that technology and morality, or technology and ethics, are intimately related.

That’s because all technologies reflect and enable human powers, choices and values.

So we have to reject the artificial, damaging split between technology and society. Doing technology right is no different to doing society right. Technology does not live outside our social world; it’s interwoven.

I want to figure out, using a blend of data-driven and humanistic tools, what are the forms of expertise, technological and moral, that can design and manage systems and ways of living that work better for people, to build better futures.

It’s not the tools themselves that can build those better futures, it’s people and the moral and social intelligence they use.

To help achieve that, we need to reunite forms of expertise currently cleaved off from each other in universities and encouraged to develop in a relationship of antagonism.

At present, we see ethicists telling technologists where they have gone wrong, or technologists telling ethicists that they are deluded or irrelevant.

How can we intervene earlier, create something which isn’t a battle between technology and ethics but something truly collaborative?

I want to take energy and desire out there and give them a path to action at the Centre – to provide for people with a desire to use their technical and moral intelligence together, and bring them into their work more.

How can we use data and AI in socially and politically constructive ways to build systems and institutions that actually support people?

The digital environments we have built are not conducive to the kind of community, democratic structures and types of leadership most of us want for our futures. We have to address their systemic harms, such as disinformation on a huge scale, which corrodes the social virtue of honesty as respect for truth.

What digital environments, platforms, processes and systems do we need to enable a future worth wanting, where we can flourish together?

Humanity needs to make progress in step with technological progress – and the clock is ticking ever faster – to make those transformational changes to the way technology and other elements of society interact.

Science and technology should be unleashing human opportunities at every turn to allow us to be sustainable and flourishing. Yet for many people on this planet, their opportunities to create new and better ways of life are shrinking thanks to political and environmental destruction. That’s a fundamental crisis.

We have to move quickly to use our technological and moral intelligence to remove obstacles to a sustainable and flourishing future. The window of opportunity is here, but will not be open indefinitely.Professor Shannon Vallor is Baillie Gifford Chair of Ethics in Data and AI at the University of Edinburgh

For more information please visit their website