When a government minister turns up to speak at your event only days after being re-elected as an MP and reconfirmed in his position, it tells you something about how important a subject is perceived to be. When that subject is artificial intelligence (AI) and its impact on jobs, you can see why he made the effort.
“We are determied that the UK can be a world leader in AI and be the best place to start and grow a digital business,” was what Rt Hon Matt Hancock MP, Minister for Digital, told the audience at CogX last month. “We need to understand the interaction between government, society and AI. Every aspect of our lives will be affected by it, so we need to get it right.”
Ethical discussion and a consideration of the social impact of technology are all too often missing from conference agendas, so it was refreshing to hear such a view coming from the very top. What Hancock came to say was not about resisting change, but ensuring that AI ends up benefitting as many people as possible. Remarkably, he even referenced the concept of a National Living Wage, where everybody receives a basic income whether they are in work or not.
“Some say that government should get out of the way and allow industry to develop it, others say that the risks are too great to do that. I reject those extremes. In every area of life I believe there is a framework of good, decent behaviour that will allow fantastic, disruptive innovation and new industries to arise, such as fintech,” said Hancock.
Considering the effect of AI on employment could not be more timely. In a report published today by the Sutton Trust, up to 15 million jobs are forecast to be lost as a result of techological disruption. It outlines the potentially bleak view of, “an elite high-skilled group dominating the higher echelon of society and a lower-skilled, low-income group with limited prospects of upward mobility and an irremediably broken social ladder”.
Alternative scenarios at opposite ends of the scale for their impact were suggested by Calum Chace, author of “The Economic Singularity” and “Surviving AI”. He said: “The best likely outcome is the ‘Star Trek economy’, where machines produce all the goods and services we need for a fulfilling life for free. A darker scenario is fascism. If all professional drivers are laid off because of automated vehicles, for example, we could start to see that or social breakdown. Let’s not do that.”
“It is seriously possible that one quarter to one half of people will be unemployable."
If you consider that over 4 million people in the US are employed in transportation and logistics, you can see what the risks are. Economic gains from automated, “lights out” warehousing and self-driving vehicles could force drivers out of work. With Google, Apple and Uber all testing this technology, that is an event sitting on the near horizon. Cutting out an entire class of labour in one step is precisely the sort of thing that drives voters towards populist and even extremist political figures.
“It is seriously possible that one quarter to one half of people will be unemployable. That has serious implications. AI can be a good thing, but not if we blunder into it,” said Chance. His own view tends towards the positive, asking, “what would a good version of the economy look like if one quarter to one half of people are unemployed, but still living fulfilled lives?”
Technology disruptions are not new - the most recent wave was in the 1980s when mainframe computers hit business and led to the automation of many tasks formerly performed by humans, from car manufacturing to book-keeping. Chance pointed out that, “the lights didn’t go out”, but he also made the point, “cognitive automation is different from mechanical automation.”
Kathryn Parson, co-founder and co-CEO of Decoded, shared these concerns. “Many people are scared by technology’s current impact on their lives. We don’t want that fear to take hold before we have even got to AI,” she said. Parson pointed out that 80% of busineses in the US are not harnessing the full effects of technology right now, even before they attempt to embrace leading-edge solutions like AI.
“What worries me is the shortfall of skills in the UK,” she added. Trying to skill up the population, especially in STEM, has support from both government and industry. But according to Parsons, some of that may have been just talk. “Three years ago, we created a data education product expecting a hunger for it, but we killed it because it didn’t happen. Now, we are getting demand from one business for 30,000 people to be upskilled in data, data science and machine learning.”
Playing catch-up is never easy, especially where the skills are complex and challenging to learn, which is very true of data and analytics. That is why the rise of the machines could have potentially challenging outcomes for humans. “Having a job is part of people’s sense of worth. Not having one causes dissent and unhappiness. Will there be quality jobs for 90% of the population? That doesn’t seem plausible,” suggested Seán Ó Heigeartaigh, executive director, Cambridge Centre for the Study of Existential Risk.
”We can’t say AI is going to change the world and that it will only deliver good things."
He does see a positive aspect in the creation of new job types and also in deploying AI to fix things which humans have screwed up, such as climate change. AI can optimise energy production and consumption in ways that, if applied across the US, might limit the global temperature rise. Said Ó Heigeartaigh: ”We can’t say AI is going to change the world and that it will only deliver good things - that won’t happen without guidance. If we asked AI to solve problems in the way we have - we have screwed up the planet. So don’t rush into it.”
As well as managing the machines, humans will also need careful handling, especially those who are displaced from employment. That is the sort of issue that governments need to consider long-term. In that respect, Hancock himself is ideally placed, as he revealed in a family anecdote.
He said: “There is a problem with technology jobs and the skills gap. I view that argument in terms of history, for example, the Luddites. The leader in Nottingham in 1810 was an ancestor of mine, Richard Hancock. He led a gang of 1,000 to smash looms. He was eventually stopped in Mansfield by the King’s Guard and transported to Australia. He had it half right - when there is a big disruption, you need to take care of those being disrupted.“