Data sharing and data within digital literacy were among the subjects addressed by expert witnesses during the second House of Lords select committee hearing on artificial intelligence (AI).
Henrietta Moore, director of the Institute for Global Prosperity and chair of culture, philosophy and design at UCL, said: “Overwhelming value in our world will be in data. Software and hardware will be held by a small number of organisations.”
Joseph Reger, chief technology officer at Fujitsu, said: “Consumers will need to give up a lot of their privacy for it [data] to work well – and it is not clear that it is a good deal for the consumer.”
Olly Buston, CEO and founder of Future Advocacy, said: “AI is supremely good at handling data. We need a new deal on data. The farcical social dance where you tick the box for consent and then quite a lot is built up on this, is a fragile thing. It needs to be fixed. We need to move to a situate where people give their data for consent for a specific function, so we may need tech to help this. Blockchain may be helpful.”
Mark Taylor, global strategy and research director at Dyson, told the committee: “How data is used is incredibly important. Personal ownership of data may be something we want to move towards as a society.”
Taylor also suggested that the decision-making process in AI should be open. “Transparency of algorithms and data are vital for public trust,” he added.
But while the tech giants may try to capitalise on the social data of consumers, Fujitsu’s Reger predicted that industrial firms will use their own datasets to build AI and machine learning. “I don’t believe the future will be a single dataset controlled by one company,” he said. “It is not going to happen because it is not in the interest of any company.”
Paul Clarke, chief technology officer at Ocado, argued that data needs to be shared. “If everyone keeps the data for themselves, we won’t create the richness from the intersections of the datasets, which is where some of the most exciting things will happen,” he said.
As Computer Weekly has reported previously, education was among the topics discussed in the first session of the select committee’s AI inquiry.
Ocado’s Clarke argued that education should focus on data skills, namely digital literacy over computer programming. “Data is the food of AI,” he said. “Teaching our children to have mastery of it is incredibly important.”
Clarke predicted that programming will be among the skills that AI will automate and said: “We need to rethink the whole curriculum. Many things we teach now will be as disrupted, just as the encyclopedia has been due to the internet. Teaching our kids to code is like tinkering at the margins because those skills will be obsolete very quickly.”
Moore added: “We seem to be doing remarkably little to teach pupils about ethics and privacy and security, and we have not thought of how to teach children about life-long learning. We will have to make the boundary between work and school very porous.”
This is an area where Clarke felt the government’s strategy on apprenticeships was actually having a negative impact, discouraging businesses to invest in their people. “We are going to have to fuzz this boundary between education and work,” he said. “We need to do more to incentivise industry towards the process of continual learning. The Apprenticeship Levy has carved a hole in the available budget that companies have to spend on continuous learning.”
Clarke argued that the Apprenticeship Levy should be transformed into a training levy, which could be used by companies in a much more holistic way.
New social contract
This would help to support life-long learning – one of the themes picked up by the other expert witnesses. Moore said: “We have neglected to invest in people for a very long period of time. We need to think about a new social contract for the future.”
She expected people to have multiple jobs, rather than a single source of income. “Multiplicity of tasks is likely to be the future of jobs,” she said. “How do we train people to do multiple tasks?”
In today’s economy, said Moore, such jobs represent the low-end of the jobs market, but this will become a problem for everyone in the future.
Future Advocacy’s Burton warned: “The path of least resistance is greater unemployment and communities will be damaged.”
Moore added: “Machines should improve productivity and create more productivity. The question is, what we do with the wealth we create?” She called for the notation of work and leisure to be readdressed.
The challenge for society is that automation and AI are unlikely to create more jobs for people, according to Richard Susskind, IT adviser to the Lord Chief Justice of England and Wales. He told the committee: “Machines are becoming more capable, and are taking on jobs that humans do. We have not seen any evidence that the new jobs being created cannot be done by machines.”
Some members of the select committee were concerned that blue-collar workers would be disrupted by AI and automation, just as in the industrial revolution of the 19th century, but the expert witnesses argued that it was white-collar workers who would find themselves out of work.
Susskind said society needs to reassess what is important. As an example, he discussed how neurosurgeons were excited about the prospect of using AI in their work. “In the end, people don’t want surgeons – they want health,” he said. “The technology allows us to think of entirely new approaches to work. It enables us to delivery outcomes in entirely new ways.”
For instance, he said, in Japan, which has a shortage of nurses, a robotic nurse is able to sense when a patient is distressed and can play soothing music or speak empathetically. “Is that better than not having anything at all?” said Susskind. “Even with an Amazon Echo, one establishes some kind of relationship.”