After seeing the recent Parliamentary report on Artificial Intelligence in the UK (more on that here) we found the AI Sector Deal a little disappointing.
It is of course good to see commitment to developing the sector, and especially the broad plans to spread developments efforts across the country and to address practical issues like the need for more PhD places and Tier 1 visas. (We're looking out for more detail on these, especially the structure of investment around the country.) But an overall vision will need to go deeper than this and set a framework for long term development.
One area that is not really addressed at all is the ethical dimension. Renowned physicist Stephen Hawking warned against the development of extreme forms of AI that could eventually take the place of human intelligence and become independent, self-improving systems.
The Parliamentary Report engages with the ethical and legal issues in some detail. Obviously data collection and use, and the control of data by affected individuals, are hot issues. But that is an area already being addressed through the implementation of the EU's data privacy reforms, the GDPR. The Parliamentary report looks beyond these issues to include discussion of wider ethical risks. How will AI affect employment? Who will shoulder legal liability when something goes wrong? What steps can we take to address criminal liability for misuse and to control autonomous weaponry. For the UK to be a world leader in AI, it will need to lead the debate in these areas.
The UK has already begun to address the wider legal framework for driverless cars, with a major review under way to analyse where we are and what the future should look like.
Projects like this take time and investment. You could not take the same approach to every new technology. But we think that AI – together with robotics - is another candidate for in-depth analysis.
The AI Sector deal flags the creation of an Office for Artificial Intelligence and an AI Council which, together, will have responsibility for implementing the Sector Deal. Alongside these, the Centre for Data Ethics and Innovation announced in the Industrial Strategy White Paper will be tasked with reviewing the current governance landscape and advising Government on “ethical, safe and innovative uses of data, including AI”.
We challenge these bodies to do two things.
- on a practical level, don't take the easy course of a continued focus on what is already happening in businesses and universities in the South East of England, but take seriously the need to drive investment in other parts of the country to boost the energy and invention being shown there.
- reach beyond the practical steps, important though these are, and take an in-depth look at how we frame the technology in a way that will ensure benefit to humans, and avoid harming them.