Managing the risks of Artificial Intelligence.
Today Artificial Intelligence (AI) permeates every aspect of our lives but are the ethical considerations about how it is wielded matching up to the extraordinary leaps in the technology?
Hughes Hall Fellow, Dr Stephen Cave, reflects on the challenges of balancing ethical and societal implications with the pace with technological innovation, and how Hughes Hall is supporting this work.
A world first
Humankind has been dreaming about intelligent machines for thousands of years: even Ancient Greek mythology incorporated the idea of intelligent robots such as Talos and artificial beings. In the mid 1950s AI generated a lot of excitement, followed by repeated slumps and hypes in interest until the early 21st century, when reality began to match imagination. Today AI is transforming every sector of society and the economy, from Fitbits to spaceships and from mobiles to medical robots. The UK government recently released its AI strategy, seeking to incorporate AI into almost every sector. Institutions and organisations including the military, the law, education and the NHS, as well as international tech giants such as DeepMind, Facebook and Google, are all involved with shaping our futures with these technologies. It is imperative that they do so responsibly.
It could be argued that the present transformation of society is on a par with the Industrial Revolution. Looking back, broadly speaking, it was the global north that benefited but even there industrialisation could be held responsible for enormous political, regional and social upheavals. The extent of many of the long-term negative consequences for society and the environment are only just being realised. As the world catapults into this next major ‘revolution’, how do we ensure from the start that AI is used for the benefit of people and the planet, and aligns with fundamental ethical considerations in both the medium and long term?
The new Master of Studies in AI Ethics & Society, supported by Hughes Hall, is designed to square up to that challenge.
Staying ahead of the curve
There is a great deal of research and teaching surrounding the development and use of AI, but the Cambridge MSt in AI Ethics & Society is the world’s first degree in its ethical and societal implications. This is not a technical degree or ‘how to do’ machine learning; the focus is on the ‘hard questions’ and seldom-aired impacts. A two-year, part-time course, it is aimed very much at already working people who aspire to be leaders in this technology.
‘As educators on this course we have two different challenges,’ explains Dr Stephen Cave (1996, Philosophy) Fellow of Hughes Hall, and Director of the Leverhulme Centre for the Future of Intelligence. ‘We have a mature, diverse, international student cohort from a rich mix of working backgrounds including business, policy, technology, law and communications. We even have a professor in bioethics on the course! Consequently they are all at different starting points. Some may have developed AI systems, but haven’t written an essay in decades; they may be deeply involved in the development of technology but less experienced in assessing its ethical impact. In contrast, the lawyers might have less experience of the technology but can think through systematic ways to regulate it.’
‘Our challenge is to create content so that this extraordinary mix will equip all graduates with a coherent and consistent skills and knowledge base. Personally, I have found the first term of this course really exciting: it’s great to be in the room and see the synergies. This isn’t just about top-down teaching; we are all learning from each other.’
‘The second challenge is that this is a rapidly changing field where there is no established canon and many controversies. The course is at the cutting edge and will keep evolving to ensure its graduates will have the skills to stay ahead as things change.’
The first cohort
When the course was announced there was worldwide press coverage. The original intention was to accept 25 students but the quality and quantity of applicants was so high that 44 were enrolled. This resulted in an outstanding group of students drawn from a surprising range of sectors and countries. It is hoped that in the future bursaries can be established to attract an even more diverse pool of applicants. Despite travel restrictions, most students turned up in person for the first residential week at Hughes Hall in September, although a few did Zoom in from Australia and the US.
‘There was an incredible atmosphere and very positive feedback from the students,’ Stephen says. ‘It is brilliant working with people, some of whom are already leading in this field, and we are building a fantastic learning network.’
Hughes Hall was keen to host the residential parts of the course. The college is the logical home for this innovative degree, as it has positioned itself as a conduit between academia and society – just like this MSt. In addition, recognising the growing importance of AI and its implications, this is one of the core focus areas of the Bridge initiative at Hughes, supported by the President, Dr Anthony Freeling, who has a strong interest in this area.
The expectation is that the MSt in AI Ethics and Society will develop leaders who can understand the complex socio-technical systems underpinning AI opportunities and challenges, and confidently tackle the hard AI questions facing their workplaces. These include privacy, surveillance, justice, fairness, algorithmic bias, misinformation, microtargeting, Big Data, responsible innovation and data governance.
‘Our hope is that those making important appointments in AI will soon be looking for candidates with our degree, who have acquired the expertise to assess and monitor the societal impact of AI in crucial systems,’ Stephen explains. ‘Looking ahead, we will work with what will become our alumni network on how they are using what they have learned and develop case studies to make it even more relevant. Our students are all working throughout the two-year course and so it should have an impact in the external world early on.’
‘From Hollywood, we are all familiar with the idea of AI rising up against humanity but there are so many risks far more immediate and dangerous than a “robot revolt”. Our ambition is that our pioneering programme will create champions of a values-driven ethical AI so that this amazing technology is used responsibly and as a force for good.’
For further information about the MSt in AI Ethics & Society, please visit: http://lcfi.ac.uk/master-ai-ethics/. Applications for the course starting in October 2022 are now open. The closing date is 31 March 2022.
This article was published in our College magazine in January 2022 – you can read Hughes issue 33 online.