Meeting of the Parliament

Meeting date: Wednesday, April 18, 2018

Official Report 821KB pdf

---

Artificial Intelligence

The final item of business is a members’ business debate on motion S5M-10161, in the name of Kenneth Gibson, on artificial intelligence. [Laughter.] It is nothing personal, Mr Gibson. The motion is entitled “Artificial Intelligence: Future Prosperity, a Threat to Employment or Existential Threat?” The debate will be concluded without any question being put.

Motion debated,

That the Parliament considers that artificial intelligence (AI) represents a potential asset to Scotland’s economy and could contribute to growth in productivity and gross domestic product (GDP); acknowledges the findings of the PwC report, The economic impact of artificial intelligence on the UK economy, which was published in June 2017 and states that the impact of AI across Scotland’s economy could boost annual GDP by up to £16,700 million by 2030; realises that the development of AI will require new industries to be formed in order to supply and service new automated solutions, therefore contributing to net employment growth; appreciates that Scotland embracing AI technology could bring benefits across society, including greater prosperity, productivity and more individual leisure time, not least to the people and communities of Cunninghame North; also notes the Cities Outlook report, published in January 2018, which suggests that 230,000 jobs in Scottish cities could be lost to automation and globalisation by 2030; is aware however, that some experts in the field of AI fear that its development will ultimately be detrimental to humanity, a matter often raised in science fiction; understands that Professor Kevin Warwick, of Coventry University, attests that networked AI systems cannot be just “switched off” when they go rogue, which is a particular problem in military applications where AI is currently being developed; understands that the Tesla car maker, Elon Musk, asserts AI to be as big a threat to humanity as climate change or nuclear war; believes that, in California, a “singularity”, which will be an ultra-intelligent machine that can make itself even more clever will have been developed within three or four decades; notes the view that, to ensure systems are developed in a responsible and controlled way, open up economic opportunities and minimise potential threats, a debate on AI and its implications for Scotland is long overdue; concludes that AI and its growing importance is an issue of global significance, and notes calls for the Parliament to address this.

17:04  

I thank colleagues who took the time to sign my motion—in particular, Gordon Lindhurst and Tavish Scott, because without their cross-party support the debate could not have taken place. I also thank Mark Dames, who is the head of public affairs at BT Scotland, his colleague Dr Andrew Starkey and Heriot-Watt University for their excellent briefings.

Just four weeks ago, the world lost Stephen Hawking, one of our most inspiring and high-profile scientists. As someone who relied upon automation and artificial intelligence to continue leading his incredible life far beyond the two-year prognosis that he was given in 1963, Mr Hawking was also one of the loudest voices warning against the dangers that are posed by the future relationship between humanity and AI. Speaking in 2014, he went so far as to say that the development of artificial intelligence

“could spell the end of the human race.”

He is not alone in having expressed such concerns. Tesla car maker and space pioneer Elon Musk has suggested that Al is as big a threat to humanity as climate change or nuclear war. Sundar Pinchar, who is the chief executive of Google, has said that the impact of artificial intelligence will be more profound than that of electricity or fire.

Although some people prefer to consign Al to the fringes of science fiction and simply ignore the inevitable universal adoption of automation, I believe that it presents perhaps the biggest challenge that society will face in our lifetime. It is therefore disappointing that the Scottish Government has not brought to the chamber a debate on the topic.

Of course, the apocalyptic notion that computers with superior intellects will eventually go rogue and turn against us is—I am assured by Al engineers—“highly unlikely”, as technical limitations hold back the ability of computers to process the same volume of information as our brains process with ease daily—even mine, Presiding Officer. Although they have potential for evolutionary and exponential growth, they may think in completely different and more peaceful ways than humans, who have survived and evolved through millennia of war, famine and disease.

What we must prepare for is the rapid acceleration of the three main technological trends that will impact on Scotland’s economy. First, there will be rapidly increasing and diversifying capabilities of machines, and data-driven decision making. Secondly, there will be a departure from traditional business models, with new start-ups trending towards asset-light and digital-platform based business. Thirdly, there will be global connectivity that will enable collaboration in decentralised online communities.

Those trends will dramatically shift the standard relationship between humans and machines by substantially reducing the involvement of workers in everyday business processes and customer transactions. Although that may be good news for businesses that benefit from streamlined processes, fears are mounting that Al technology will destroy jobs and, indeed, entire industries faster than it creates them, thus creating mass unemployment and handing market control to a handful of dominant firms that are quick to harness the new technology.

The current wave of technological change is so far reaching that it has been described as the “fourth industrial revolution” in an excellent report that was published by the Scottish Council for Development and Industry in collaboration with BT Scotland, entitled “Automatic ... For the people?” Of course, each preceding industrial revolution has produced winners and losers, but the distinction here is that the influence and effects of Al and machine-learning technology will be ubiquitous, transformational over a few short years and not reserved to a few sectors. Unless we radically reassess our workforce—especially those in the most vulnerable jobs—we risk considerable social dislocation.

As the motion highlights, the recent report by the Centre for Cities suggests that as many as 230,000 Scottish jobs could be lost in our four biggest cities over the next decade. However, according to the SCDI and BT report, as many as 837,290 jobs are at high risk of being lost to automation, from 8 per cent in education to 44 per cent in retail and 63 per cent in water supply. The last alone would mean 10,642 fewer jobs.

Highly skilled private sector occupations are expected to increase, while lower-skilled and more routine activities will shrink. We are not just talking about the future. Businesses of all types already use Al to forecast demand, hire staff and provide customer services. In 2017 alone, companies globally spent £15 billion on Al-related mergers and acquisitions—more than 26 times more than in 2015, which demonstrates the momentum that Al now has. The McKinsey Global Institute estimates that applying Al technology in marketing, sales and supply-chain departments could be worth £2 trillion in profits and savings over the next 20 years.

In financial services, Al already shapes new processes in financial controls, regulatory reporting, applicant checks and referencing data, thereby eliminating human error on critical financial reporting. In healthcare, Al can eliminate subjectivity in patient diagnosis and use algorithms to connect symptoms and test results, thereby delivering more accurate prognoses. Its uses also range from detecting criminal activity to identifying web material that is designed to radicalise Facebook or YouTube users.

Al could transform the workplace and give employers unprecedented control over staff. From Amazon-style wristbands that track the efficiency of warehouse staff to smart ID badges that track interaction between employees, data will be harvested and used in ways that we may not yet conceive. Big brother could be watching you.

The picture looks a little gloomy on the jobs front, but does Kenneth Gibson accept that, in hospitality and coffee shops, for example, a lot of people want to be served by a person and that that makes a real difference?

Yes. I certainly hope that human interaction remains at the forefront. I am one of those people who never uses a machine at Tesco or Asda, for example; I always prefer to be served by a shop worker. However, the trend is quite simple and straightforward and is heading in one direction. We have to adapt our economy and consider the fact that some people who own coffee shops might agree that some people will want to be served by a human, but other owners will look just at the bottom line.

It seems reasonable that Al could be used to screen for anomalies or to flag up differences in pay between genders and races that conscious or unconscious bias could cause a human to overlook. However, as the Cambridge Analytica saga has demonstrated, data is a valuable asset, and our laws are not yet fit to protect workers from automated surveillance that goes beyond the consent that is baked into employment contracts.

I am not here to provoke alarm or to theorise about the end of days, but rather to encourage the Scottish Government to join other Governments that have, in collaboration with industry and civil society, already set out their Al and automation strategies. Indeed, the programme for government states the intention to transform Scotland into a nation that will lead in Al, machine learning, data analytics and low-carbon energy. Nevertheless, I doubt that Al is currently being given the high priority that is required.

Germany already has a 10 to 15 year strategy to advance the adoption of new digital technologies across industry, and federal departments are exploring aspects of Al, including the ethics of self-driving cars—as was raised in Ivan McKee’s members’ business debate—the impact on the workplace, and use of drone technology. France has commissioned a national Al strategy, and Estonia is exploring the use of automation in healthcare, finance and other sectors.

With some of the world’s leading research universities being in Scotland—notably Heriot-Watt University—already undertaking cutting-edge work, and a plethora of data-intensive businesses having chosen to set up in our cities, I am confident that with the right strategy and outlook Scotland can make the most of the opportunities that are afforded by those innovations.

The Fraser of Allander institute has already advised that much of the research that has been undertaken into the potential impacts of technological change on Scotland has used United Kingdom-wide data and applied it to Scotland’s unique industrial structure. To predict more accurately and plan for technological transformation, the Scottish Government should lead the way in researching what tasks and activities will be impacted by automation, and the distinct impact that that will have on Scotland’s businesses and workers.

With careful planning and proper regulation, technological change will create growth and help businesses to grapple with a shrinking working-age population and weak productivity growth. According to the PricewaterhouseCoopers report entitled “The economic impact of artificial intelligence on the UK economy”, which was published last June, the impact of Al across Scotland’s economy could boost annual gross domestic product by £16.7 billion by 2030 through developing new industries.

I have talked at length about technology, but what really matters is Scotland’s people. It is critical that education and training equip not just our young people with the skills that are necessary to adapt to upcoming technological changes, but that they do so for the 80 per cent of Scotland’s current workforce who will still be of working age by 2030. There is consensus that the principles and design of the curriculum for excellence are right for the opportunities and challenges of life and work in the 21st century, but there is more that we can do to prepare our workforce better for the economy of the future.

We are all familiar with the stereotype of the worker edging closer to retirement who suddenly finds that there is no market for their skills and is unable to adapt to new technology. However, in this digital era, people could face that prospect far earlier in their careers. The Scottish Government must foster a culture in which lifelong on-the-job learning is not just an optional extra, but an inherent feature of working life.

I am sure that everyone has their own vision of Scotland’s future, but surely we must all agree that artificial intelligence is an issue of global significance that cannot be ignored. In the words of the chief executive of Centre for Cities, Andrew Carter,

“The time to act is now”.

17:13  

I did not realise that the words

“Big Brother could be watching you”

would be the cue for John Mason to make an intervention, so I will not use them. I see that he is leaving the chamber at this point anyway.

I thank Kenneth Gibson for bringing this debate to Parliament. I signed his motion because it raises a number of important points about artificial intelligence that are worth reflecting on.

As the convener of the Economy, Jobs and Fair Work Committee, I can see that there are potentially huge and positive implications for Scotland in the coming years. As has already been highlighted, other issues may need to be looked at more closely. However, it is important not to be alarmist about future developments. I think about the concerns when computers first became mainstream not that long ago; people were worried that everyone would be out of a job. We soon learned that—at that stage anyway—a digital copy and a hard copy of everything had to be kept, which meant that we had a double workload. We may be beyond that in this paperless Parliament, but that is certainly how it was in the beginning.

New jobs were created by the computer industry, and we are all familiar with the resulting cyberworld, whether we want to be or not. Research by Deloitte has found that, although technology is estimated to have cost 800,000 lower-skilled jobs between 2001 and 2015, 3.5 million higher-skilled jobs have been created in their place as a result of technology.

We should not exaggerate the effects of Al. After I had signed the member’s motion, a constituent wrote to me to express concern about it, suggesting that it may be an

“exercise in futurism reminiscent of the predictions made in the 1960’s that we would now be having holidays on the moon”.

I do not think that any of us have had a holiday on the moon—at least not yet today.

Nevertheless, my constituent agreed that a debate on Al was well overdue in the Scottish Parliament. Their principal concern was that that debate should be an evidence-based one. I hope that we can all agree on that, because there are identified and legitimate concerns, such as the social and economic implications of increased automation; the use of obsolete data; the protection of personal privacy; and the inappropriate application of biases and prejudices established from real-world data transferred into automated systems without adjustment. Those are just a few of the legitimate concerns about the advance of AI.

However, let us remember those concerns at the same time as noting that, as it says in the motion, the effects on productivity and the resultant contribution in boosting the economy can be positive and immense. Increased productivity is something that we could do with in Scotland, given the lack of growth over the past eight years. We can harness our advantage by the progress that we have made on the AI front, which could prove key to being at the forefront of the technology of the future.

That aspect is recognised in the UK Government’s industrial strategy, in which

“growing the Artificial Intelligence and data driven economy”

is one of the four “grand challenges” that the UK can take advantage of. The autumn budget included £75 million on Al-related developments last year and £21 million for tech specialisms in the UK, including for a hub here in Edinburgh, which is a city with a successful Al track record—and, indeed, in this very chamber.

We must recognise the benefits that Al brings at the same time as we remain live to the risks and the difficulties it could also create.

17:18  

I thank Kenny Gibson for bringing this debate to Parliament. A few weeks back, we discussed driverless cars in a members’ business debate sponsored by Ivan McKee. Mr Gibson made a particularly interesting speech, in which he referred to the trolley dilemma and wondered how a driverless car would decide to interact in such a situation. That speaks to some of the problems and the challenges that we will have with AI; it also speaks to the profundity of our attempting to understand the impacts that it will have.

In looking forward, it might be useful to begin by looking back and situating what might happen within the broader context of previous industrial revolutions. Mr Lindhurst referred to the potential for alarmism and overstating what might happen. If we consider the impact of the first industrial revolution, we see that steam and the railways led to some of the most profound changes since the agricultural revolution. We can also consider the profound impact of synthetics, including dyes, chemicals and plastics. However, one of the most profound inventions was the washing machine. It emancipated many—by a large majority, women—from domestic drudgery—

Men, too.

And men, too.

Do you think?

The microwave oven, the kitchen, plumbed water and sanitation have had the most profound impacts. If we are to suggest that a fourth industrial revolution as a consequence of AI and automation will be equally profound, we must consider what we mean by that.

I think that AI could have an impact on that scale. The reason for that is that, if we reach the stage at which machines can start to learn for themselves, the potential will exist for exponential growth. Some people forecast that machines will pass the Turing test within our lifetime. When that happens, it will give rise to a range of considerations that seem to belong to the realms of science fiction but which could become real. If a machine develops the capacity to think and, potentially, to feel, should it have rights and responsibilities and duties and obligations? In future, we might face such questions. The fact that that is even a possibility suggests how profound the impact could be.

Would the right to vote be one of the rights that would have to be considered in the scenario that Mr Arthur postulates?

Such a scenario might sound so speculative as to be almost farcical, but it becomes less so if we think about the role that machines could play in law as paralegals and the introduction of machines to generate automated responses for the civil service, which some jurisdictions are experimenting with. Machines could have a role in supporting politicians in doing the job of representing their constituents. For example, a machine could potentially take on a piece of constituency casework. Therefore, there are significant implications for our democratic system and how we think about it, even before we get to the stage of having genuine artificial intelligence.

Speculative concerns aside, the most profound point concerns the potential threat to jobs. As has been highlighted, the issue is whether we are talking about job displacement or job replacement. I welcome the Scottish Government’s latest publication on the subject, “Technological Change and the Scottish Labour Market”, which I am sure the minister will refer to. It takes a very balanced view. My experience of the Economy, Jobs and Fair Work Committee’s evidence taking is that although there are people on the fringes who make predictions that we face a catastrophe or, alternatively, that artificial intelligence will make no difference at all, the general consensus seems to be that, as with previous industrial revolutions—for example, when the use of horses as a means of transportation and power declined—new jobs will emerge. That is potentially the situation that we will be in.

It is extremely important that we are cognisant of the opportunities and risks that exist. Although we are talking about a speculative and inchoate development, we have a duty as politicians to put such ideas into the public domain and to make sure that the population at large is aware of them, because there is an inevitability to the whole process. When the changes in question take effect, it is vital that we—and, most importantly, the public—are prepared.

17:23  

As others have done, I thank Kenny Gibson for bringing the debate to Parliament. For some strange reason, some of my colleagues—and, indeed, some of Mr Gibson’s—burst into laughter when I said that the debate was on artificial intelligence and that it would be led by Kenny Gibson. Who knows why they would do that? Frankly, I find that shocking.

Tom Arthur struggled to make the point about women being liberated from the kitchen through the invention of the washing machine, plumbing and various other things. Far be it from me to point out that men were liberated through such inventions, too, and far be it from me to say that artificial intelligence might liberate women from men completely; that is a novelty for another time.

The Labour Party has, of course, long been committed to protecting workers’ rights, ensuring high standards of working conditions and creating the opportunity for organisations and businesses to thrive. I have no doubt that, in the 21st century, the world is changing, the economy is changing and work is changing. Equally, I have no doubt that the on-going life-changing technological advances that we are seeing will change the face of work as we know it.

Given that we are on the brink of the next industrial revolution, it would be foolish of us to approach such a fundamental change to our country’s industrial landscape with anything less than the enthusiasm that previous progress was met with. The opportunity to innovate our sectors, to improve the experience of workers and to strengthen our position on the world stage through the likes of artificial intelligence should be embraced with open arms, while ensuring that precautions are taken to minimise any negative impacts that might arise.

It is vital that we seize this change and maximise its potential benefits, but we need to stay in control. We should be shaping how automation works for us, rather than allowing artificial intelligence to shape us.

That means working with the trade unions and working alongside employers to dictate how best artificial intelligence can fit into our economy to guarantee that we get the most that we can from such progress.

The changes that automation presents are far greater than we previously thought. Automation will affect every part of our economy. However, I am sure that Kenneth Gibson agrees that the current state of our economy leaves little room for complacency. That was evidenced in the recent figures on the minimal growth that Scotland’s economy achieved in the previous quarter. That suggests that if we want to improve economic performance in Scotland, as we all do, automation could provide an opportunity to drive a significant boost to our productivity and our gross domestic product.

It is worth repeating that, as Kenny Gibson highlighted, in June 2017 PricewaterhouseCoopers told us that the impact of AI across Scotland’s economy could be to boost annual GDP by up to £16 billion.

Will Jackie Baillie give way?

Indeed.

I am grateful—

I am sorry for not calling you, Mr Arthur, but I was waiting with bated breath to hear what you were going to say next. I call Tom Arthur.

Thank you very much, Presiding Officer. I would be interested to hear Jackie Baillie’s view on the potential for a combination of the internet of things, big data and artificial intelligence to revive the idea of a planned economy.

We believe that planning in the economy is critical. Take for example some of the recent debates that we have had on procurement. I think that our view on the opportunity to secure more of the supply chain in Scotland is shared across the chamber. If we can use artificial intelligence and big data to achieve even more of that, I do not see what the problem with that is. I want us to get the maximum that we can for our investment. If artificial intelligence helps with that, we should embrace it.

I want to see better conditions for workers and maximum productivity in our sectors, but we know that industries such as transport, retail and administration are likely to diminish in size. Industries such as those currently hold the majority of Britain’s 900,000 zero-hours contracts and Scotland’s something like 75,000 such contracts. In many cases, those are for low-skilled, low-wage jobs.

Labour has put a huge amount of time and effort into trying to change the exploitative nature of those jobs. I commend to the chamber our industrial strategy and Tom Watson’s future of work commission, in which there are plans to ensure that workers will receive the retraining required to take full advantage of the high-skilled, high-wage jobs that often come hand in hand with automation.

We are seeing new technology and telecommunications industries emerge and information and communication technology and digital tech jobs in Scotland are expected to increase from 84,000 to 150,000 by 2020. That is an increase of 11,000 new skilled jobs each year in that sector. However, we have a challenge: there is a distinct lack of skilled young workers and we need to train even more 16 to 24-year-olds to meet that challenge.

I see you waving at me, Presiding Officer. Despite the earlier hilarity, I think that artificial intelligence provides us with an opportunity to secure economic prosperity for future generations. There is an opportunity to make it work for our economy.

17:29  

I thank my colleague Kenny Gibson for bringing this very important subject to the chamber for debate this evening. The subject of AI, or artificial intelligence—I am not sure whether Jackie Baillie had a different AI in mind in the earlier part of her speech—is one of the biggest challenges facing our society and economy at the moment. It is a hugely broad subject and I will focus on two particular areas. First, there is the economic impact and how we manage that. Then I will say something briefly about the moral impact that was mentioned earlier in the debate.

The scale of the economic impact has been referenced by the reports that we have talked about: the city impact report, the PWC report and the SCDI automation report. The PWC report talks about a potential growth in the Scottish economy due to AI of 8.4 per cent by the year 2030. It refers to that of the UK, which it thinks could grow by 10.3 per cent due to the difference in the structure of the economies, which we need to bear in mind as we talk about how best to exploit the opportunities of AI for the Scottish economy.

As many members have said, we have been here before. I remember that when I was growing up in the 1970s, people talked about the changes that technology was going to make. In the early 1980s, we had huge unemployment as a consequence, but we came through that and new technologies and new jobs took up the slack. We do not still have huge pools of people typing out letters; they have been replaced by technology. We do not still have a million people working in coal mines, which are largely gone, and such jobs have been replaced by others. AI and technological developments have huge potential dramatically to increase productivity and enable us to do a lot more that is, as Kenny Gibson mentioned in relation to healthcare, of better quality.

The question is what we, as a society, should do to prepare to manage that best. It is about focusing not on the breadth of what might happen but about picking out a number of specific sectors in which we can leverage the skills base that we have and put some investment and focus into them. We should look at the research and development strengths that we have and work with academia, business, Government, trade unions and the third sector to identify a handful of sectors that we should invest in and focus on with a view to becoming world class. I have previously brought the subject of self-drive vehicles to the chamber for debate. That could be one of the sectors; we need to identify another.

In business, one of the hardest things that we need to do as far as strategy is concerned is to anticipate when to move on from a very successful business model and build a completely different one for the future. We need to embrace disruptive technology now, before it is too late. We must ask how we configure our education and skills system to be able to deal with it, and how we create the attitude that a job—or even a career—is not for life and that we must constantly reskill into different jobs. My moving into a career in politics at the age of 50 is perhaps an example of that. As Jackie Baillie said, the important thing is that we should stay in control through all that and be able to manage the impacts.

I want to say a brief word on societal impact. Clearly, a lot of things will happen in such a transition and it will be very difficult for individuals and their families. At this stage, it is worth talking briefly about the role that a universal citizens basic income might play in smoothing out the transition. It would give people a support network and also the confidence to be able to take risks, identify opportunities and start up businesses. Even if they fail, they will know that there is a support network there to enable them to move from one career to another without hitting huge financial hardship as a consequence. That has to figure very largely in where we are going with AI.

I will make a very brief point on the moral side of things. Clearly, the singularity concept is potentially scary if things go wrong. Tom Arthur entertainingly asked whether machines should have rights and responsibilities, which is perhaps something for discussion—who knows?

In conclusion, I would like to ask the Government to focus on four or five areas. The first is what we are doing to identify the specific sectors on which we should focus to take best advantage of the coming technological revolution. What are we doing on education and skills to prepare ourselves for that? What are we doing to ensure that the social transition is as smooth as possible? That is perhaps where the citizens income could come in. We need to start a debate on the moral aspects of that and, as Kenny Gibson identified, perhaps to pull together to identify an AI strategy that allows us to move forward with some confidence.

17:34  

I, too, congratulate Kenny Gibson on securing time for the chamber to debate what is a hugely interesting topic, on which I am delighted to be able to contribute. It gives us so much scope for what we can talk about.

Death, taxes and change are the three guarantees in life. Of course, change is happening at such a rate that we increasingly have to adapt to it and change our skill sets, as Ivan McKee pointed out, just to stay in the job market or even keep up with life. We humans are instinctively wary of change and tend to resist it, but, as all the Trekkies in the chamber will know, resistance is futile.

Technology has advanced at an incredible rate in my lifetime. I have mentioned in the chamber previously that I did not have a mobile phone until I was in my 30s and I remember black-and-white television, when we had to get off the sofa to change only three channels. Can you imagine that, Presiding Officer? Now, we are on a technological highway that moves at such a speed that it is increasingly difficult to keep up with it.

I have always been interested in the mythical technological singularity, which comes from the prediction that there will be a point in time when machines will be smarter than human beings. Ray Kurtzweil, Google’s director of engineering and a well-known futurist, predicts that we will hit that point within the next 30 years or so; in fact, he reckons that it will happen around 2029. We should take heed of him, because he has an 86 per cent accuracy rate for the 147 predictions that he has made since the 1990s. He says that the singularity will lead to

“computers having human intelligence, our putting them inside our brains, connecting them to the cloud, expanding who we are. Today, that’s not just a future scenario. It’s here, in part, and it’s going to accelerate.”

That takes me back to Seven of Nine and the Borg.

We all have to accept that the singularity will come sooner or later, but the question is whether we should fear it. Everyone knows, of course, that when machines become smarter than humans, they tend to take over the world, matrix style. Right? As Kenny Gibson indicated, many of the world’s science and technology leaders, such as the late Stephen Hawking, Elon Musk and even Bill Gates, have warned us about that kind of future. The issue certainly helps to keep the world’s sci-fi film studios busy and that, in turn, keeps us royally entertained.

Kurtzweil suggests, though, that the singularity—the point when a single brilliant AI enslaves humanity—is just fiction. I suggest that AI offers us opportunity. We might consider the idea of a cybernetic society to be more of a fantasy than a glimpse into the future, but there are people with computers in their brains today: Parkinson’s patients. That is an example of cybernetics getting a foothold. Perhaps technology will be invented in future that can go inside our brain and help our memory, which would have implications for dementia sufferers, for example

Perhaps the vision of machines taking over the world at the point of the singularity should be replaced with the vision of a future of human-machine synthesis. That would, literally, open up whole new worlds in terms of space exploration. How can we as a species, with our frail bodies and minds, travel the vast distances across the galaxy that are required for us to continue in our thirst for knowledge and our need to consume resources? Currently, we could not survive journeys in space to explore other parts of the solar system, let alone the stars beyond. AI is the most feasible option that science has come up with in that regard, and perhaps it will even involve downloading our own consciousness into a machine. Currently, a Mars rover continues to send back information from the surface of Mars. Ultimately, we will need to leave this planet if the human race is to survive, so perhaps we need to rethink our definition of what constitutes a human being.

As Kenny Gibson’s motion suggests, we do not know what is coming down the track, but we need to ensure that Scotland is ready to take advantage of the opportunities that AI will undoubtedly bring. We have a great track record in developing new technologies in Scotland. However, whatever our thoughts and fears about AI are, we can confidently say that AI is not so much a case of “I’ll be back”, but more a case of being here to stay. We need to embrace the opportunities that it will bring.

17:38  

Thank you, Presiding Officer, for letting me speak in the debate. I had not intended to do so, but I am, because I have been fascinated by the contributions so far.

In the 1980s, I was studying for my computing degree. It was quite an interesting degree at what is now Glasgow Caledonian University because, although it was a science degree, we also studied psychology and business and accounting. As part of our psychology course, we were asked to look into the effect that the computing industry might have on future generations and what impact it could have on working lives. That was put to us in the context of the riots during the Wapping dispute, which arose because some newspaper production was moving from the manual printing process to a digital one. We saw what happened at Wapping because of that and how the police reacted to it, and it was a really interesting lesson.

I remember reading an essay—I cannot remember who wrote it—that talked about the human race as a whole, our collective psyche, and how we respond to technology. The author talked about Copernicus and his view that the sun rather than the earth was the centre of the solar system, which was of course proved years later with Foucault’s pendulum. Galileo was excommunicated from the Catholic church, so shocking was the thought that human beings were not the centre of the universe.

The author suggested that the next stage that would have a shocking effect on humans would be when artificial intelligence came about. It would have the same effect on the human race because we would no longer be the sentient being in our universe. It was fascinating to read that, especially in the context of the real ways in which technology was changing. We talked about the Luddites and how they approached technology in their time, destroying weaving machines because they were a threat, and we came right up to the modern day and what was happening in our country at the time.

That whole thought about how we approach technology has remained with me, and if it has taught me anything, it is that the people who stand against the advancement of technology very rarely win. It is something that we cannot hold back, and the way to get an advantage from that is to be the leaders and the experts: to be the people who lead in new technology and innovations.

As someone who worked in IT, I want to be clear that a lot of what we have talked about today is not AI but what, in my time studying, was called expert systems. That is about taking the knowledge that we have as humans, applying it to a computing function of some kind and getting a result from that. The computer does not do anything other than replicate what it has been told to do by humans. That is absolutely not artificial intelligence; it is just about capturing data and using it in a positive way to achieve an outcome.

In the health situations that I looked at, information was captured from medical people to arrive at a diagnosis based on the steps that those people would go through to achieve a diagnosis. When we look at current opportunities such as the work that is being done at CENSIS on sensor technology, again, that is all about capturing information and environments and using that in a positive way. We do not have real artificial intelligence yet.

The warnings that we have heard should be noted and we should be cognisant of them. We have examples from Marvin the paranoid android, and Holly in “Red Dwarf”, right through to HAL 9000, who killed off his entire crew. The warnings and concerns exist, and although we can look at ASIMO, the Honda robot, which is cute and looks benign and unthreatening, we know that the technology in there can be used in a military form to weaponise. As with all such things, the most important thing is that we understand the technology and use it for the benefit of humanity.

I call Paul Wheelhouse to respond to the debate for about seven minutes—although, given the way it has gone, I would say that you can have as long as you like, really, minister. [Laughter.]

17:43  

Thank you, Presiding Officer. I, too, congratulate Kenneth Gibson on securing the debate, and I welcome the speeches that we have heard from members throughout the chamber. Regardless of which study we talk about—whether it is the Centre for Cities one, the BT and SCDI one or the PWC one—and which we rely on for our estimates of the impacts in society, we cannot overstate the significance of the issue for our economy, Scotland’s people and our workforce, so it is right that Kenneth Gibson has brought the topic to the chamber. I take on board his point about Government time and I will play that back to my colleagues, but I reassure him and members throughout the chamber that we take the issue particularly seriously.

The topic of artificial intelligence and other emerging technologies has always interested me. A number of references to science fiction have been made today, and I mention on the record “Do Androids Dream of Electric Sheep?”, “Blade Runner” and its successor film as personal favourites. They raise some ethical issues about how robots and artificial intelligence can be used, and that ties in with what Clare Adamson has just said. The important point, which cut across all the speeches in the debate, is that the issue is not just the technology’s development but whether it can be used to benefit mankind and our planet, rather than do us harm.

The topic that we are debating interests me and resonates strongly with the Scottish Government. It is an area to which attention is increasingly being turned. Tom Arthur talked about the work that Jamie Hepburn is leading in respect of technological change in the Scottish labour market, which relates to some of Ivan McKee’s points about the need to consider the labour market and adjust the curriculum in our schools, colleges and universities to ensure that our young people are prepared for the world that they will encounter.

In March last year, my colleague Keith Brown, Cabinet Secretary for Economy, Jobs and Fair Work, was asked to provide a welcoming speech at the European robotics forum. It was the first time that that prestigious event had been held in the UK. The forum attracted more than 800 delegates and provided a valuable opportunity for policy makers and stakeholders to engage.

The organisers’ choice of Edinburgh as the location for the conference is a testament to Scotland’s strengths in computer science research and proof that our skills and expertise in the area have achieved recognition across Europe. Indeed, the Edinburgh centre for robotics, which is led by Heriot-Watt University and the University of Edinburgh, is a UK-wide collaborative body, and, as Brian Whittle said, the University of Edinburgh is working with NASA on its Valkyrie robots, which are to be used in future missions to Mars.

We are at the cutting edge of the technology and we can be proud of that, but Kenneth Gibson rightly challenged us to think about how it will impact on the people of Scotland. It is understandable that there has been much focus on the impact on occupations and jobs. There is sometimes an emphasis on the loss of jobs and employment, but technology might replace tasks and make processes easier, with the jobs remaining but the way in which they are done changing fundamentally.

Will the minister comment on the metrics that we use for the economy, such as headline employment figures, GDP and productivity, which often do not capture exploitative work, zero-hours contracts and work that is not stimulating? We are seeing the hollowing out of medium-skilled jobs, and although there might be growth in high-skilled jobs, greater automation could lead to an increase in low-skilled jobs. Does the minister agree that we need a superior set of metrics if we are to understand what is happening and direct change so that it improves people’s wellbeing and overall quality of life, instead of simply focusing on increased GDP?

I certainly do. The Government has been considering, with stakeholders, alternative measures of the economic success of our society, which go beyond GDP. I entirely take the member’s point about the current measurements’ difficulty in picking out issues such as zero-hours contracts and other exploitative practices. AI will further complicate our ability to understand the impact on individuals and translate that into wages and wage growth. We need to be mindful of the issue.

Emerging technologies such as AI present exciting opportunities for Scotland, as Jackie Baillie rightly said, but we must acknowledge that with those opportunities come a number of concerns. To be fair, Jackie Baillie recognised that.

It is a part of human nature to have concerns about the unknown, but where there is an unknown, we can learn—and learning is an area in which Scotland thrives. Our universities are considered to be world class and have a history of excellence in fields such as data science, machine learning and artificial intelligence. I take Clare Adamson’s point that we need to be careful about how we use the term “artificial intelligence” and I defer to her knowledge, as an information technology specialist, of the difference between expert systems and true artificial intelligence.

Emerging technologies can drive growth and productivity, as Gordon Lindhurst said. As we move further into the modern commercial environment, our industries are required to continuously adapt. Ivan McKee asked about the sectors on which we are focusing. Manufacturing is clearly an area to which we will have to give a lot of attention. Equally, in financial services, for example, the growth of fintech is already causing concern to many members. The loss of branches across the country is partly a response to the move to technology, which will increase as artificial intelligence is used more. These things are happening now.

Technology is bringing opportunities in the public sector. In healthcare, for example, there are opportunities to improve patient experience and quality of life. Stephen Hawking’s points about the threats were well made, but he clearly benefited from technology, and we need to identify opportunities to improve quality of life for individuals—as I said, we need to focus on areas in which there can be a gain for mankind.

By integrating processes such as automation, we can remain competitive in the global marketplace. That is particularly key to the future of our manufacturing base, which is a sector that has been highlighted in a number of studies as one that will be strongly impacted on by emerging technologies.

That is why the Scottish Government has committed £48 million to a national manufacturing institute for Scotland at Inchinnan to help accelerate innovation by enabling manufacturing companies to trial and test new processes, applications and technologies. We are also supporting development through the manufacturing 4.0 service to help companies to understand how emerging technologies can be integrated into their businesses effectively and efficiently. That service will be launched properly very soon. The NMIS will also help to support our workforce by providing resource to develop and enhance the skills that they and employers need, resulting in more competitive businesses while safeguarding jobs.

We are a small nation, but we are proud to be a vibrant, inclusive and outward-looking digital nation. The Scottish Government has a vision for making the most of data by championing across Scotland a trustworthy use of it for public benefit. Delivering innovation using our skills in data science and artificial intelligence techniques is an important strand of us achieving that vision and we are working to accelerate that through data-driven research.

Scotland’s refreshed digital strategy, “Realising Scotland’s Full Potential in a Digital World”, which was published in March last year, sets out plans for ensuring that we put digital at the heart of everything we do. Data innovation plays an important role within that strategy and, along with digital, it will create an irresistible force to drive innovation in our public services.

There is also the importance of transparency. This work needs to be carried out under robust ethical and governance frameworks. Kenneth Gibson, Clare Adamson and others made powerful remarks about the need for ethics, and Ivan McKee’s earlier debate on autonomous driving also touched on these issues.

We are investing £300 million in the Edinburgh and south-east Scotland city region, including £60 million for innovation. That investment, and investment from the UK Government, which we should acknowledge, will help to secure our place as the data capital of Europe and to create an environment that will nurture and attract further innovation and investment to Scotland.

I have much that I could say about cyber-resilience, but I am conscious that I have already run out of time. This has been a valuable debate and I know that members on all sides are focused on ensuring that, in Scotland, we take an ethical and informed approach when considering artificial intelligence. I note the concerns that Mr Gibson raised, and it is important that Parliament acknowledges that concern and is mindful of the potential impact on the workforce of Scotland.

I hope that I have assured members that we are taking steps to strike the correct balance when considering the needs of economic development against our social and ethical values. The future will bring many opportunities, and I hope that we all agree that Scotland is well placed to be a global leader in the development of artificial intelligence and other emerging technologies.

Meeting closed at 17:52.