Artificial Intelligence and the Labour Market

Lisa Cameron Excerpts

Wednesday 26th April 2023 Westminster Hall

[Mick Whitley (Birkenhead) (Lab)]

I beg to move,

That this House has considered the potential impact of artificial intelligence on the labour market.

It is a privilege to serve under your chairmanship, Dame Maria. I am grateful to all hon. Friends and Members who have taken the time to participate in this important debate. It is a particular pleasure to see my hon. Friend the Member for Ellesmere Port and Neston (Justin Madders) in his place. I wish to draw the attention of the House to my entry in the Register of Members’ Financial Interests.

The rapid advance of artificial intelligence technology poses a severe threat to the labour market and to workers’ rights. The negative effect of AI on the workforce cannot be ignored, as it has the potential to displace jobs, lead to economic inequality and erode the rights of workers. AI has the capability to automate jobs and various industries, which could result in widespread unemployment and exacerbate existing socioeconomic disparities. Low-skilled workers, who are already vulnerable to exploitation, are likely to be the most impacted, leading to a growing divide between the haves and the have-nots.

Furthermore, the implementation of AI in the workplace could result in the violation of workers’ rights such as privacy, autonomy and fair pay. The use of AI to monitor and control workers could lead to increased exploitation, discrimination and the creation of a toxic work environment. If left unchecked, the rise of AI could lead to a future where workers are replaced by machines, and human dignity is sacrificed for the sake of corporate profits. The deployment of AI in the workplace must be accompanied by strong regulations and policies that prioritise the wellbeing and rights of workers.

Governments and companies must take responsibility for the harmful impact of AI on the labour market and take immediate action to prevent its negative effects. Failure to do so would result in an irreparable loss of jobs, economic inequality and a violation of workers’ basic rights.

For Members who have heard me speak before in this House, that introduction must have felt unusually stilted, or perhaps uncharacteristically eloquent. That is because it was written entirely by ChatGPT—one of a number of increasingly sophisticated AI chatbots that have become readily accessible in the past few months. At this point, let me reassure my parliamentary researcher, who is watching this debate, that he does not need to worry about his P45—yet. The unusual distinction of being the first Member of Parliament to openly read AI-generated text into Hansard belongs to the hon. Member for Bosworth (Dr Evans). Like him, I have chosen to turn to one of the most widely used AI-powered technologies to illustrate the rapid advances taking place in the field of artificial intelligence, and the potentially devastating consequences that this could have on workers in every sector of the economy.

Of course, the impacts of this AI revolution will be felt far beyond the labour market. Information is an increasingly valuable commodity; it is also a potential weapon of war. The danger is simple: technologies such as ChatGPT and DALL-E could be used to proliferate dangerous misinformation and subvert our already compromised democracy. We need further and extensive scrutiny of the risks and of the steps that we need to take to better protect our constituents’ data privacy.

I have chosen to use the limited time available today to look at the impact of artificial intelligence on the labour market, and particularly on workers’ rights. That is not only because I have spent my adult life fighting for workers’ rights, but because it is in the labour market that that change is happening most rapidly, and it is in the everyday experience of work that the disruption of AI is being most keenly felt.

We have heard much in recent years about how we stand on the edge of a fourth industrial revolution. That revolution is now well under way; its effects will be profound and far-reaching. Every part of our public life will be transformed. I want to be clear: I am no enemy of progress. We should embrace the potential of AI to change our lives for the better, whether by improving diagnosis and treatment of disease or by driving sustainable economic growth that can benefit us all. Just as the first industrial revolution brought about an era of unprecedented wealth for an elite few but condemned the British working class and colonised people across the world to a life of precarity and poverty, the AI revolution will create again—if we allow it to do so—a world of winners and losers.

[Dr Lisa Cameron (East Kilbride, Strathaven and Lesmahagow) (SNP)]

I thank the hon. Gentleman for making an impressive speech and extremely good points about the welfare of workers. As a union rep, I agree that we must have safeguards around AI developments. Does he agree that in order to make this new technology available to all, we should seek to level up across the UK and ensure that coding opportunities and the jobs of the future are available to young people in all areas, including deprived areas?

[Mick Whitley]

The hon. Member makes a good point. When it comes to AI, all workers need protections.

Research by PricewaterhouseCoopers suggests that AI will be responsible for 46% of the UK’s long-term output growth. It promises job creation in sectors such as health, education, and science and technology. At the same time, it threatens devastating job losses in sectors such as manufacturing, transport and public administration. Some 7% of all UK jobs could be automated away within the next five years, and as many as 30% could disappear within 20 years.

The last time we experienced systemic economic displacement on anything like that scale was during the deindustrialisation of the 1980s and 1990s. The architects of that policy believed that nothing should be done to support those communities that carried the cost of the economic and social fallout, the legacy of which my constituency of Birkenhead continues to live with to this day. They followed the ancient mantra that the strong do what they can and the weak suffer what they must. We must not repeat that mistake again. I have called today’s debate to make an urgent plea for a rights-based and people-focused approach to artificial intelligence, and for a process that puts the voices and interests of workers at its heart. In this new machine age, we must assert more than ever the fundamental right of all people to a basic level of economic security and dignity at work.

--- Later in debate ---

[Dean Russell (Watford) (Con)]

It is a privilege to speak in this debate, and I thank the hon. Member for Birkenhead (Mick Whitley) for securing it. I wanted to apply for it myself—he beat me to the chase, which is a wonderful thing.

Before I became an MP, one of my final clients was in the AI space. It dealt with artificial intelligence and psychology—I believe that my first entry in the Register of Members’ Financial Interests was my final bit of work for it—so I have seen this technology evolve over many years. We often talk about technology revolutions, but this has been an incredibly fast evolution.

We are seeing Moore’s law, which related to the size and scale of technology, affect society. The scale of what is happening right now is both inspirationally amazing and terrifying at the same time. It will absolutely shape the job market and the type of jobs that come through over the next few years. It will shape how people interface with their co-workers, with technology, with society and with politicians. It will affect every aspect of our lives.

I am particularly concerned about the use of artificial intelligence for deception. I have long said—not necessarily in the Chamber, so I put it on the record now—that there should be in law something that I would call the Turing clause. It would mean that when technology is used to deceive somebody into believing that they are talking to a real person or engaging with a real business, whether for entertainment or for any other purpose—for instance watching a deepfake, which is perhaps for entertainment purposes—it must be crystal clear to them that they are being deceived.

I will give some examples. I was recently speaking to somebody who works in the entertainment industry, running studios where they record sound, voiceovers and music. They said—I should declare that I do not know the scale of this issue and have not looked into the numbers—that lot of the studios are often being used to record voiceovers for AI companies, so that the AI can learn how to speak like a real person. We all know about fraud and scams in which somebody gets phoned up from a call centre and told, “Your insurance is up,” or by someone pretending to be from the Government. We saw, awfully, during the covid crisis how those horrible people would try to scam people. Doing that requires a number of people in a space.

Now imagine that AI can pretend to be somebody we know—a family member, for instance—and imitate their voice. It could call up and say, “I need some money now, because I am in trouble,” or, “I need some support.” Or it could say, “This is somebody from the Government; your tax affairs are an issue—send your details now.” There are a whole load of things going on in society that we will not know about until it is too late. That is why a Turing clause is absolutely essential, so that we are ahead of the curve on deception, deepfakes and areas where technology will be used to fool.

One incredibly important area in relation to the labour market that is not often talked about is the role of AI in creativity. DALL-E 2 is one of the tools, and there are many others popping up now. They can create artwork and videos almost at the speed of thought—typing in a particular phrase will create amazingly beautiful pictures—but they are pooling those from places where real artists and real musicians, with particular styles, have contributed. That is then presented as AI creativity. That could kill the graphic design industry. It could prevent people who are in the early stages of life as an artist, in both the visual and music worlds, from ever having an opportunity to be successful.

Just recently, Drake and the Weeknd—if I have those artists correct—had a song that was put online. I think that it even went on Spotify, but it was definitely on some streaming services. Everybody thought, “Gosh, this is a fantastic new collaboration.” It was not. It was AI pretending to be both of those artists with a brand new song. Artificial intelligence had created it. It was not until after the fact, and after the song had been streamed hundreds of thousands of times, that the big music companies said, “Hang on—that isn’t real. We need to stop this.” Then it was stopped.

In the case of social media, it took us many years to get to the fantastic Online Safety Bill. I was very fortunate to be on the Draft Online Safety Bill Joint Committee. Its Chair, my hon. Friend the Member for Folkestone and Hythe (Damian Collins), is in the room today, and he did a fabulous job. Getting to that point took 10 or 15 years. We do not have 10 or 15 months to legislate on AI. We probably do not have 10 or 15 weeks, given where we will be in a matter of days, with the new announcements and tools that are coming out.

[Dr Cameron]

I thank the hon. Gentleman for making those extremely important points. Just last week, we had the Children’s Parliament at the all-party parliamentary group on the metaverse and web 3.0. The children were excited about the opportunities of AI and the metaverse, and we were told on the day that the World Economic Forum predicts that technology will create 97 million new jobs by 2025 alone. But like the hon. Gentleman, they were also very concerned about what is real and what is not, and they were concerned about the mental health impact of spending much of the day in an altered reality setting. Does the hon. Gentleman agree that we need much more research into the mental health impact on staff and young people who are engaged in AI?

[Dean Russell]

I thank the hon. Member for her comments. Mental health is a passion of mine—I had a ten-minute rule Bill about ensuring that mental health first aiders are in the workplace—and I agree wholeheartedly. We saw that in evidence given to the Draft Online Safety Bill Joint Committee; Rio Ferdinand talked, including in his documentary, about the fact that what is said online can affect a person’s real life. The challenge with artificial intelligence is that it will not just be able to say those things; it will probably know precisely how to do the most harm, how to hit the right triggers to make people buy things and how to fool and deceive people to ensure they hand over money or their rights.

I will move on because I am conscious of time. I know we have quite a long time for this debate, but I do not intend to use it all; I promise. I think that the creativity part is absolutely essential. A few weeks ago, I predicted in Parliament that, in the next year or so, a No. 1 song will be created by artificial intelligence for the first time. I have no doubt that a No. 1 bestselling book will be written by artificial intelligence. I have no doubt that new songs in the voices of artists who are no longer around, such as Elvis Presley, will be released, and that actors who are sadly no longer alive will play starring roles in new films. We are seeing this already on a soft scale, but it is going to become more and more pervasive.

It is not all negative. I do not want to be a doomsayer. There are great opportunities: Britain—this wonderful country—could be the home of identifying and delivering transparency within those industries. We could be the country that creates the technology and the platforms to identify where artificial intelligence is being used; it could flag up when things are not real. It could, for example, force organisations to say who they are, what they are doing and whether they have used artificial intelligence. I think that will create a whole new world of labour markets and industries that will stem from this country and create all the jobs that we talked about earlier.

I am also concerned that we do not often talk in the same breath about artificial intelligence and robotics. In the industrial world, such as in warehouses and so on, there has been a rise in the use of robotics to replace real people. Office jobs are changing due to artificial intelligence. The role of accountants, of back-office staff and of both blue and white-collar workers will change.

As was stated earlier, the challenge with robotics is on things such as defence. Artificial intelligence is being used in robotics to get way ahead of the scale of where we are now. We really need to take that seriously. ChatGPT was probed. People tried to catch it out on different aspects of its response. When asked how it would steal the nuclear codes, it outlined how it would do it. I am not trying to give any bad actors out there any ideas, but it explained how it would use AI to control drones, and how they would be able to go in and do certain things. Hopefully, it got it all wrong. However, if AI is in not just our computers and mobile phones, but in drones and new robots that are incredibly sophisticated, incredibly small and not always identifiable, we need to be really wary.

There are many positives, such as for detection in the health sector and for identifying things such as breast cancer. Recently, I have seen lots of work about how artificial intelligence could be layered on the human aspect and insight, which was mentioned earlier, and enable the identification of things that we would not normally be able to see.

There is huge positive scope for using data. I have said previously that, if we were to donate our health data to live clinical trials in a way that was legitimate and pseudonymised, artificial intelligence could be used to identify a cure for cancer and for diseases that have affected our society for many centuries. In the same way that it has found new ways of playing chess, it might find new ways of changing and saving lives. There is great opportunity there.

Many years ago, I wrote an article called, “Me, Myself and AI”. In it, I commented on areas where AI is dangerous, but I also mentioned opportunities for positives. I would like to make one final point on this: we must also make sure that the data that goes into the AI is tracked not only for things such as royalties in creative industries, but for bias. I wrote an article on that a while ago. If we take a sample, say within a health context, and take that data based on only one ethnicity or demographic, the AI will develop options and solutions for that group. If we do not have the right data, regarding diversity, going into the analysis, we risk not being able to identify future issues. For example, sickle cell disease might get missed because the data that the AI is using is based only on clinical trials with white people.

There is a wide-ranging issue about what is being fed into the systems around AI and how we ensure that we identify where AI is being used—hence my point about a Turing clause when it comes to deception. We also need to know where it is being used, including in Government. We need to look at the opportunities, too: whole new industries around how we monitor AI, apply it and use the science of it.

AI is already there in the spelling of “Great Britain”. We have a great opportunity to be ahead of the curve, and we need to be because the curve will be moving beyond us within a matter of weeks or months—and definitely within years.