Artificial Intelligence 11:00:00 Geraint Davies (in the Chair) Before we begin, I encourage Members to wear masks when they are not speaking, in line with current Government guidance and that of the House of Commons Commission. Please give each other and members of staff space when seated and when entering or leaving the room. 11:01:00 Nick Fletcher (Don Valley) (Con) I beg to move, That this House has considered the future impact of artificial intelligence on the economy and society. It is a pleasure to serve under your chairmanship, Mr Davies. I called for this debate because artificial intelligence will very soon affect every aspect of our lives. The past demonstrates that everyone will willingly allow this to happen—just look at the rise of the internet algorithm. We willingly hand over our data, enabling service providers to produce complex algorithms that sell us more products and get us to click on specific websites. The question cannot be whether we stop the rise of AI; it should be whether it can be effectively regulated. Naturally, I was pleased that the Government published their national AI strategy last month, which at least began the conversation about how we can manage this technology so that it benefits our economy, workforce and society. Let me give some examples, starting with the driverless car. We must realise that it will probably not be long before insurance companies acknowledge that fewer accidents will occur than in man-driven vehicles. Furthermore, research suggests that advances in technology will enable X-rays directed by AI to diagnose cancer far more quickly and far more accurately than the best of our consultants. While some of these technologies may seem far off, they have already taken over many unskilled low-paid jobs. After all, it was not that long ago that we ordered McDonald’s coffee in person. Then, one day we were met with a giant screen. Personally, knowing full well the implications of that over time, I deliberately went to the counter and ordered my coffee in person to protect people’s jobs. I did so until one day when the counter was not manned and a nice lady stood next to the giant iPad and said, “Come on. Use this.” Now, every time I go to McDonald’s, I use the giant screen. The nice lady has gone. That is the crux of the issue. AI technology is often introduced to aid the pre-existing workforce. Yet, just like McDonald’s, managers eventually realise that their workforce can be replaced wholesale, and the AI technology is what is left—doing what humans were doing, but doing it better. Let us take another example: mowing the lawn. While many people find gardening a chore, our desire to keep pristine gardens means that the gardening and landscape business can employ 160,000 people. Yet, as those people retire they are likely to be replaced by artificial intelligence technology as it becomes more capable, because employers will not be liable to provide sickness pay or holidays. AI can cut grass; how long before it can cut hedges and pick soft fruits? Throughout covid, we have seen the classroom change too. Am I saying that we should remove the teacher? Of course not, but with the rise of AI will we always need teaching assistants, administrative staff or examination boards? I do not know the answer, but it is essential that we start asking these questions. Jim Shannon (Strangford) (DUP) It is important to have a debate on this subject and Westminster Hall is a great place to have it. UK GDP could be up to 10.3% higher by 2030 because of artificial intelligence and its impacts from consumption-side product enhancements, and more importantly as a result of widening consumer choice and making available more affordable bespoke goods. Does the hon. Gentleman agree that it is time that the Government, especially the Minister, backed the AI brokers by ensuring that there is a skilled workforce? There is a workforce to be formed out of AI, and that is what we should be focusing on. Nick Fletcher I thank the hon. Member for that intervention and I will come on to exactly those points later. I do not think that a red wall Conservative can ever make a speech without mentioning Brexit and trade deals. In light of Brexit, AI will probably be more utilised than ever before to move goods across borders. In warfare, too, the rise of drones to maintain and expand our geopolitical influence is already apparent, yet drone technology is already being used in combination with AI and we see armies across the world, from France to Russia, using AI-controlled drones in conflicts. Let me reflect briefly on the Departments that could be affected by the examples that I have given: Transport; Health; Work and Pensions; the Department for Business, Energy and Industrial Strategy; Education; the Department for Environment, Food and Rural Affairs; the Department for International Trade; and the Treasury. It is quite a list. So, what is the problem? Well, it is what I want to highlight today, so that, as parliamentarians and as a Government, we can start to have a frank and honest discussion on this issue. First and perhaps most worryingly, the rise of AI technology is likely to decimate people’s jobs. I have heard it said, “Well, people have threatened that before,” and looking at the unemployment figures we see that Britain had some of the lowest unemployment figures ever before the pandemic. However, this new AI revolution will be different from the industrial revolution when it comes to employment. As a result of the massive expansion of AI in many sectors, AI will affect many people’s lives and pretty much every job sector. AI will infiltrate everything, everywhere. And just as with internet algorithms, we will all be willing participants. Will it happen overnight? No. It will take time. As I have already alluded to, there has been increasing use of AI for many years. However, the gradual rise of this technology means that policy makers, the Government and the public are not aware of its creeping challenges. Little by little, we as a society are becoming more dependent on it, and little by little it is making life’s many tasks more manageable. So, which jobs will be affected and—more importantly —when? Let us start with jobs in call centres and fast food restaurants, as well as driving jobs, which, yes, means every taxi driver, every delivery driver and every HGV driver. In total, that amounts to over 600,000 people. Warehouse workers, shop assistants, postal workers, parking attendants—the figure for all those jobs is over 3 million people. If that was it and the list did not expand further into security, education, health and defence, I am confident that a forward-thinking Conservative Government could manage such economic stresses. Yet even when we are discussing the jobs that are most at risk, we must remember that employees in such jobs are often younger people, so our young people’s future is most at risk. One of my biggest beliefs is that the devil makes work for idle hands and the worst idle hands are young ones. A young person with no job often believes that they have no value. Although that is not true by any stretch of the imagination, we cannot have an entirely new generation of young people thinking it about themselves. I am sure that many people who are interested in this subject are fully aware of the game Go and the experiment to see whether AlphaGo, an AI programme, could beat a renowned Go player, Lee Sedol. For those people who do not know, Go is apparently one of the hardest games in the world to play, with an almost infinite number of moves and, most importantly, no real patterns for AI to follow. Consequently, many people were amazed that AI won the first three games. Personally, I never doubted that it would. Yet what struck me was Lee Sodol’s reaction. When he lost the third game, it seemed to destroy him. To me, he looked empty—a person with no value. Now, if an educated man of that calibre can be made to feel worthless because of the abilities of an AI programme, we can imagine what future generations trying to enter the workforce may feel like as a result of AI. Lee Sodol won the fourth game, which brought his pride back, despite the fact that he then lost the fifth game. It should also be mentioned that AlphaGo went on to beat every other competitor while playing them all at the same time. To me, that proves not only the infinite capability of AI, but the damage that it may do to individuals, unless we look ahead and consider ways in which we can use AI while still keeping people feeling valued within our workforce. That brings me to the national AI strategy. Pillar 1 of the strategy, “Investing in the long-term needs of the AI ecosystem,” talks extensively about the lack of AI skills within the economy. An AI Council survey found that only 18% of 413 respondents from the fields of academia, business and the public sector believed that there was sufficient provision of training and development in AI skills for our workforce. Clearly, there is an AI skills shortage. However, while the strategy mentions the “Skills for Jobs” White Paper, published in January this year, and states that it will work to ensure businesses have the necessary skills to utilise AI technology through the skills value chain, it offers little in acknowledging the huge problem before us. The industries I have mentioned employ nearly 4 million people, most of whose jobs are to be made essentially redundant in the coming years. A reskilling scheme from the Department for Education, here and there, will not tackle the issue at hand and ensure that millions of individuals, many of whom we currently consider skilled workers, will not become unemployed over the course of this century. If we have learned anything about the levelling-up agenda, it is that people in places such as Don Valley want to have jobs that provide value and meaning to their lives. Let there be no mistake: unless we sufficiently equip huge numbers of our workforce over the coming years, many will never secure work, let alone skilled, meaningful work. Getting this right is key to the Government maximising the impact of their levelling-up agenda. A good start may be establishing a new college in Don Valley that specialises in coding. I say to the Minister, let our economic revolution begin in Doncaster. That is the employment issue covered, but if AI touches every aspect of our lives, then why not the democratic process itself? We recognise that individuals, myself especially, are not infallible. MPs make mistakes, and Governments too. People sometimes initially vote against what seems to be their own interest, yet despite this we accept that democracy is the best form of Government or, as Churchill said, “democracy is the worst form of government—except for all the others”. If we accept that AI can apparently make more efficient decisions, what role does that leave for MPs and even for the ordinary voter? AI challenges how the state should be run, what the public wants or what piece of AI technology it believes is most efficient. I welcome that the Government have committed to a strategy to work with international partners on shaping international norms and standards relating to AI, which puts the shared values of freedom, fairness and democracy at the heart of the development of this technology. Can the Minister let the House know exactly how democracy will be underpinned through such work? Can he inform those listening to the debate what work is currently being done with the Foreign, Commonwealth and Development Office to come to an internationally binding agreement that ensures artificial intelligence will not be developed in a way that will subvert the democratic process? It is important that we begin these discussions now, because the rise of AI is inevitable. We cannot stop it and history teaches us that every move to oppose the rise of technology is doomed. That is why we had an industrial revolution, while the Luddites became a mere footnote. AI is already here and both hostile countries and our allies are using it. Therefore, we must engage with this technology if we are to maintain our position as a leading world economy. It is fantastic news that the Government have begun to think about balancing regulation with innovation in the national AI strategy, yet we need to do much more if we are to avoid facing severe societal and economic destruction as a result of this emerging new technology. I finish by stating that I know this debate is difficult for the Minister. I will be the first to admit that I do not have the answers to the problems posed, but I look forward to working with the Minister and stakeholders in order to fully utilise the benefits of AI, as well as mitigate its inevitable effects. The Minister will no doubt agree that it is important for us to recover from covid, with the entire global economy in mind. In the light of that, we must work with our international partners, just as we have done to beat covid, to work on regulation of AI so our democratic way of life is preserved. In conclusion, when it comes to AI, I first ask the Minister and the Department to continue to embrace this technology and work with businesses to ensure there is adequate research and development investment in this industry. Secondly, I stress to the Minister how important it is to integrate AI within the levelling-up agenda. The Government should therefore plan ahead so that young people in places such as Don Valley get the technical skills at college to build artificial intelligence programmes in the future. Lastly, while embracing AI, the Government should be wary of its potential to cause disruption within society and should mitigate any negative effects of this emerging technology. 11:15:00 The Parliamentary Under-Secretary of State for Business, Energy and Industrial Strategy (George Freeman) It is a great pleasure to serve under your chairmanship, Mr Davies; to be back on the Front Bench to make the case for science and technology in this country; and to respond to my hon. Friend the Member for Don Valley (Nick Fletcher), who has done his constituency and constituents a service by raising these important issues, and in exactly the spirit of our late and lamented colleague, Sir David Amess. We need in this place constituency MPs who speak for the fears, worries, anxieties and concerns of their constituencies, as my hon. Friend eloquently has. I hope to address some, if not all, of the points he made. I reassure him that they were well made and well heard and are important to the Government as we set out our plans for the UK to be an AI powerhouse. I am framing my new role as Minister for Science, Research and Innovation around two key projects. First is the mission to be a science superpower. In many ways we already are, but we need to maintain that to be able to grow a modern, innovative, prosperous and high-skilled economy. Secondly, crucially, is to ensure that, off the back of the pandemic, the opportunities created by Brexit and debt challenges owing to the global financial crisis and the pandemic itself, we build a much more innovative, productive, high-skilled and competitive economy by harnessing technology and innovation, to make the UK an innovation nation. Fundamental to my mission is to make sure that the benefits currently enjoyed—not only, but heavily—in the golden triangle are spread so that we can build clusters of new sectors, new jobs, new companies and new technologies all around this country. That means not only in Scotland, Northern Ireland and Wales, strengthening the Union, but in constituencies such as mine, which is not 40 miles from Cambridge but feels 100 years away, and like my hon. Friend’s, which hear of this technology revolution but do not see the opportunities on their own doorstep. I am grateful to my hon. Friend for so fluently raising these issues. Of course, we already use AI across whole rafts of our society and economy to huge public benefit. I have seen, through my own career and as the former Minister for Life Sciences, the incredible power of AI software in looking at genomic and phenotypic records and very quickly—in a way that no number of scientists on their own could—identifying opportunities for new drug discovery or targeting drugs at the right patients, which has huge benefits for patient safety. In cyber-security, AI is right on the frontline of our ability to counter some quite mischievous and dark forces, in terms of both national security and economic fraud. AI already plays a crucial role for the environment. For example, in agritech, using AI with satellite data helps to identify where to apply chemicals in isolated parts of a field; rather than spraying a whole crop or field, AI identifies, by field patterns and visual optics, where chemicals need to be applied. In fact, the use of AI in plant genomics allows us to develop a whole raft of drought and disease- resistant crops, helping sustainable development. In air traffic control, thankfully, huge computing power is applied to ensure that planes never bump into each other; it is important to have pilots when there is an emergency, but actually the AI at the heart of our electronic air traffic control system is keeping us all safe. AI is also used in other ways, including in the gaming sector, which is a huge driver of innovation and opportunities in this country, often rather below the horizon. I dare say that there is probably a cluster of games entrepreneurs in Don Valley. The gaming industry in this country is huge and drives a lot of innovation in AI that then has applications in healthcare and broader industry. My hon. Friend raises an important point about public trust and confidence. I am positive about the importance of this technology for creating opportunities and jobs but, crucially, the public must be with us, and they must have confidence in our regulatory framework. I am glad that he referred to the report of the taskforce on innovation, growth and regulatory reform, which I led with my right hon. Friend the Member for Chingford and Woodford Green (Sir Iain Duncan Smith). In that report, we argue that leaving the EU presents an opportunity for the UK not to race to the bottom but actually to race to the top: to set values-based regulation for innovation that reflects the values of the people of this country. In a whole raft of new technology sectors, the world is grappling with how to regulate: AI, autonomous vehicles, nutraceuticals, functional foods, clinical trials and digital health. We are respected internationally as a setter of standards. As my hon. Friend made clear, standards must be embedded in the values that go with the Union Jack around the world. If we can regulate with values in a way that supports innovation, I am very confident that his constituents will benefit. That goes right to the heart of the Prime Minister and the Chancellor’s historic commitment—it is the first time in my life that I have heard such a strong commitment from Conservative leaders—to end the low-wage economy that is reliant on overseas labour. The only way to do that is by harnessing innovation to create a more productive, more competitive economy. That is the way to raise the living standards of all of our constituents—my hon. Friend’s and mine. Having heard the Chancellor and the Prime Minister announce that groundbreaking commitment at party conference, I am not sure that it has yet sunk in: that the Conservative party is absolutely determined to raise the living standards of people around the country, to raise wages and to move on from a 40 or 50-year cycle of economic boom based on very cheap labour. That is good news for my hon. Friend’s constituents as well as mine. The computing revolution led to huge fears that we would see the automation of everything and mass redundancy, but in fact the UK has become a huge global software and computing power, creating hundreds of thousands of jobs. I am confident that, if we deal with the issues that my hon. Friend raised and get the regulation and skills environment right, we will similarly become a powerhouse for new AI industries. I will deal with the important points that my hon. Friend raised on skills, public trust, levelling up and ensuring that these technologies create jobs all around the country, values and security. In fact, I will go this afternoon to the Pacific Future Forum in Portsmouth to join leaders from the economies of the Association of Southeast Asian Nations. There I will highlight the UK’s commitment, through our global science superpower mission, to an international framework for the safe use of AI and to using our collective liberal democratic economic heft and values to ensure that the west is developing these technologies without inadvertently leaving ourselves open to dark forces. I will summarise where we stand and why this is such an opportunity. At the moment, the UK ranks—believe it or not—third in the world in terms of the development and deployment of AI technologies, behind only the USA and China. That is an extraordinary global advantage. AI is going to be as transformational as computing, and we are currently in bronze position in the Olympic medal table. We have a huge lead. It is important that we do not drop that lead, and that we build on it to create a prosperous economy. A third of Europe’s AI companies are here in the UK, which is twice as many as any other European country. We are also third in the world for AI investment, behind only the US and China, attracting twice as much venture capital investment into AI companies as France or Germany. We are in a very strong place in the global race to harness AI. I turn to the points my hon. Friend made on skills, because they are very important and the Government take them seriously. Since the AI sector deal that we launched in 2018, we have been making concerted efforts to improve the skills pipeline, not just to ensure that those vital high-technology skills are there for industry but to ensure that all—his constituents and mine—have an opportunity to participate in this economy. That is why we have increasingly focused on reskilling and upskilling: so that, where there is a level of displacement, there is redeployment rather than unemployment. That is why, through the Office for Artificial Intelligence and the Office for Students, we have funded 2,500 more postgraduate conversion courses. Those include courses particularly for students with a background not in science, technology, engineering or maths and students with a near-STEM background. There are also 1,000 scholarships for people from under-represented backgrounds, particularly women, black and disabled students. Those courses are available across the UK and, as my hon. Friend referenced, Sheffield Hallam University within the Sheffield city region is leading in this, and is one of the universities delivering those courses, which are hugely popular with students. I see that no Opposition Members are present, but Government Members will be pleased to remember that at the recent Conservative party conference the Chancellor announced that the programme will be doubled, creating 2,000 more scholarships. South Yorkshire is quite a powerhouse in AI, with Sheffield University. There are 16 sectors for doctoral training in AI across the country, of which Sheffield is one, training 1,000 more PhDs. There is the Sheffield centre specialising in speech and language technologies—an area where the university has long pre-eminence. Like so much of the UK, South Yorkshire is in the process of reinventing itself and its economy, and I have every confidence that it will do it as well as everywhere else, not least because of Sheffield Robotics, a leading company and employer in that region. Sheffield’s advanced manufacturing research centre currently offers more than 300 apprenticeship places to local jobseekers in the AI sector, so there is a lot to be proud of and confident of in the region. We are also seeing applications of AI at the Centre for Child Health Technology in Sheffield as part of the Olympic Legacy Park, where AI is being put to use to assist clinicians in identifying tumours via scanning. In the national AI strategy, the Government committed to supporting the National Centre for Computing Education to ensure that there is a wider reach and access to AI courses for people all around the country. My hon. Friend mentioned the importance of the Government gripping this matter strategically, and I want to reassure him on that. The Council for Science and Technology wrote to the then Prime Minister in 2013 to advise on what it called the coming age of algorithms and the need for new research to look into these matters. The Government created the Alan Turing Institute, which is now the national hub of expertise on AI and data science. Following the independent AI review in 2017, we created the Office for AI and now the independent AI Council. We also announced at the time the Centre for Data Ethics and Innovation, which is really important and goes to the heart of some of my hon. Friend’s concerns. If we are to lead in harnessing these new technologies we need to lead in regulation based on values and ethics, and reflect them as he did in his speech. I am very pleased that the Centre for Data Ethics and Innovation was a recommendation from the Royal Society and the British Academy in their separate data governance report. Earlier this year, to improve public discourse on AI the CDEI engaged widely with the public and published its findings in June. We are committed to trying to grow that conversation. It recommended that the Government develop a standard for transparency on algorithms in the public sector, which I am delighted to say is work now close to completion. We have to lead this through the public sector as well as the private. That, again, speaks to the importance of values. The international dimension is vital. I reassure my hon. Friend that in my first four weeks I have already chaired meetings with other western democracies on the importance of research security, because AI can be used for industrial espionage and intellectual property theft. It is an issue that we take very seriously, and I am jointly responsible with the Under-Secretary of State for Digital, Culture, Media and Sport, my hon. Friend the Member for Croydon South (Chris Philp), for the Office for AI, which develops a cross-Government approach. As my hon. Friend the Member for Don Valley referenced, the national AI strategy sets all that out. We have required regulators such as the Information Commissioner’s Office, the Competition and Markets Authority, the Financial Conduct Authority and Ofcom to specifically consider the risks and benefits of AI within their sectors. Earlier this year, through the CDEI and the Office for Artificial Intelligence we set out with other regulators a project to remedy skills gaps in terms of knowledge of AI in the regulatory landscape. Every regulator will need to think about how it uses AI, and the risks of AI in its sector. Internationally, we have set up the Global Partnership on AI, the first multi-lateral forum, and we co-chair the data working group. The UK is playing a leading role in international discussions on AI ethics and potential regulations, including at the Council of Europe, UNESCO and the OECD, which is partly why I am going to the Pacific Future Forum this afternoon. Time is against me, but I hope that I have addressed some of my hon. Friend’s points, and reassured him that we take them very seriously. We will harness the benefits of the technology to create those hundreds of thousands of jobs only if we bring the public with us, which we are committed to doing. Question put and agreed to. 11:29:00 Sitting suspended.