By Canon Mike D Williams
Most books about Artificial Intelligence (AI) are soon out of date. Not this one. Despite the technology speeding ahead faster than authors can write books, Verity Harding sets out fundamental principles that should apply throughout time. Her premise is that people and political processes should regulate the development and use of AI in the world today.
Verity Harding was a political advisor to Nick Clegg when he was Deputy Prime Minister. She left that job to go and work in AI. She thinks hard about technology and society.
Her argument is that humanity has previously grappled with other major changes in technology that we can draw lessons from. She investigates three, the space race, in vitro fertilization (IVF), and the internet. In each example collective agreement has been reached to avoid creating problems or to achieve improvements for society.
At the height of the Cold War in 1967 the USA and Russia cooperated in establishing a United Nations Outer Space Treaty. That treaty established key principles, boundaries and barriers as to what was acceptable for countries’ use of outer space. Nuclear weapons were banned from space and the moon kept demilitarised. Today, could the USA and China create the diplomatic conditions to achieve a convention to ensure the use of AI for the benefit of the planet? Given that it is tech companies rather than states that are driving the AI revolution, and as regulators, nation states seem unable to catch up, my view is that it is much harder than regulating outer space.
IVF led to the birth of Louise Brown in 1978 and raised multiple ethical questions. Harding tells the story of how the UK government eventually set up the Warnock Commission that engaged in wide consultation. After some years of delay the Thatcher government set up the regulator authority to oversee practice and research. The story in the USA was very different. How good a single nation example is for AI in a global industry is a question. Equally, the time it took in the UK to regulate IVF is also a concern when faced with a rapidly developing technology like AI.
The final example is of the internet, where common open protocols and collaborative organisations were established through Tim Berners Leigh, the inventor of the internet. He was determined that no company or nation state would own or run the internet (see his book ‘This Is For Everyone’). This example is not without problems, but it speaks most directly to the digital technology era.
Whilst I agree with Harding’s analysis that people and their governments should be closely involved in agreeing the principles and boundaries for the use of AI, I find her examples unconvincing. The rush to create autonomous weapon systems and the commercial drive for companies to reject regulation makes it hard to see a way forward. Yet the principle that AI should be regulated for the benefit of us all is hugely important.