Dr Robin Niblett CMG
Ladies and gentlemen, welcome to Chatham House. I’m delighted that you would come and join us today for this discussion we’re going to be having on Preparing for Digital Transformation. As I was thinking about this at the beginning, I wondered whether preparing is – sounds like a rather passive word in today’s world. Coping with might be one dimension, leveraging, obviously, would be another, as well. But, either way, we’ve got three super participants with us today and we built the event slightly around the target of opportunity of Tom Siebel being in London, at this time, and the issuing of his new book, ‘cause I know you’ve written three books before, Tom. But his new book on Digital Transformation: Survive and Thrive in an Era of Mass Extinction, and I don’t think he’s necessarily talking about the same extinction as the, kind of, rebellions that’ve been happening here, we’ve – been happening here in London. But nonetheless, it does try to capture, I think, the sense of deep transformation that the digital world is taking us through today, and so we’ll have an opportunity to talk about it in a minute.
I do want to say, before I introduce our three speakers, this event is part of our series on the Digital Society Initiative, which cuts across all of our work at Chatham House, and tries to bring together policy practitioners and digital practitioners, from their respective word – worlds, so that we can bridge some of the occasionally deaf or mute conversations that happen between those two worlds and have an opportunity, certainly for our members, all of you here, and guests, to have an opportunity, as well, to interact with these big topics that seem to be affecting so much of our world, but which we often only know small segments about.
Just to mention, this meeting is on the record. It is not under the Chatham House Rule. It is being livestreamed, so welcome to our members and others, who are joining us online. And what we’re going to do is have a bit of a conversation, here, amongst ourselves and then, you know, go ahead and engage all of you with any questions or thoughts you might have, and that you might want to raise with us on the topic. So, in terms of introductions, Tom Siebel is the Founder and CEO of C3.ai, which, I think, in itself, probably the name implies what it’s involved in, but it’s a leading provider of enterprise artificial intelligence software. And I think you’ve explained, Tom, to me, when we’ve spoken before, the specificities of that, but some people may want to dig into it. As I said, he’s the Author of the new book, copies of which are here, which you’re more than welcome to take with you at the end, I think it would be a good opportunity for you – all of us to get up to speed quickly. He’s been in the business for a while, I will simply say. But his, as having started off in Oracle, having set up Siebel Systems, it was then bought by Oracle in 2006, he set up C – C3.ai, when? It was about…?
Tom Siebel
January 2009.
Dr Robin Niblett CMG
2009? So shortly thereafter, and has really built it into the one of the key companies in this field, and we’re going to have an opportunity to talk to him about obviously, where he sees technology going. I would also mention, in the context of one of our la – my later introductions, he’s also the Founder of Siebel Energy Institute, so thinking about how this topic gets into other spaces as well.
Conor Kehoe, Advisor, former Senior Partner at McKinsey & Co. We were saying, on German Street its headquarters, they’re moving very shortly, I believe. He founded McKinsey’s private equity investor industry practice, but as originally a Software Engineer. He also has led McKinsey’s Tech and Telecom Practice, and I will say this here, in case you don’t say it later on, it says here he’s completed Stamford’s Machine Learning Course, just last year. So, you are fully back up to speed and will be able to take on some others, maybe on the panel, or in the audience, with your knowledge as well, Conor. So thank you very much for joining us.
And, last but very much not least, my Chatham House colleague, Bernice Lee, who’s the Executive Director of the Hoffmann Centre, Sustainable Resource Economy at Chatham House, one of our Research Directors here. She used to run, actually, our Energy, Environment and Resources Team, went off to the World Economic Forum and become Director of Climate Change and Resources Security over there. Has had experience of a number of other sectors, as well before, but I think will really be able to bring the policy perspective as well into this conversation, from her vantage point there.
So, off we go. Tom, if I could start with you, I mean, just as a set-up question, I suppose, tell us a little bit about what you think are the radically new elements of this element of digital transformation, ‘cause when I saw the title Digital Transformation, I – even I feel I’ve been in a digital transformation for the last ten/15 years. What is changing now, and what do you see being fundamental for the future?
Tom Siebel
I’ve been in the information technology industry for four decades, and I would say, in the last decade, particularly, we see this constant refrain in the boardrooms of Rome, Paris, London, Shanghai, New York, this mandate for digital transformation. And I couldn’t quite figure out what that was all about, and it was being driven by the CEO, as opposed to what, analogue transformation? What does this mean? And so, I gave this some thought, for about eight years, and then wrote this document, in the last couple of years, where I draw some parallels between evolutionary biology, and what’s going on in the business world.
So, if we’re looking at evolutionary biology, I mean, the planet’s been around for about six and a half billion years. We’ve had life on the planet for three and a half billion years, and the last 400 million years there have been five mass extinction events, in the last 440 million years. And in these mass extinction events, as many as 96% of the species on Earth would be eliminated, the most recent being this one with which we’re all familiar, called the K–T extinction, with the meteor hitting the Yucatán and we had massive climate change and, I think, 76% of the species on Earth were eliminated, including the dinosaurs. And when these events happened, they’d be followed by mass re-speciation events with, you know, with these organisms with new DNA would fill the vacuums filled by those who were eliminated. And, so in the case of the K–T extinction, the mammals filled the void that were formerly occupied by the dinosaurs, and that worked out pretty well for us.
Now, if you look at what’s going on in the corporate world, this digital transformation issue, really is an existential event, and we are looking at a mass extinction event, in the corporate world, in the 21st Century. In fact, in the last 18 years, 52% of the Fortune 500 companies have ceased – have – and they’re gone. You know, Westinghouse, Toys “R” Us, Siri Roblox, soon, GE, okay? And they’re just – they’re gone, they’ve been merged, they’re bankrupt, think of Lehman Brothers., and that’s at 52% in 18 years and experts expect that, in the next ten years, as many as 40% of those remain will disappear. And so, this CEO-driven mandate for digital transformation really is an existential event. And so, I argue that what is driving us, then we have now, in this vacuum, we have these companies with new DNA that you read about every day: Uber, Tesla, Amazon, Airbnb, that are filling the vacuum, and what are they all about?
They’re all about using this new generation of technology, Elastic Cloud computing, AI, big data, IOT, to solve problems in entirely different ways. Look at Uber, no cars, no drivers, and they’re upending the transportation business. Look at Amazon, you know, until recently, no storefront, okay? But they’re all about big data, AI, IOT, and they’re just rolling retailing. In the United States we had 8,000 retail outlets close last year, and so I think what is going on, is we have this new step function of technology in the form of these technology vectors that I’ve mentioned that are converging, and there’s a sense that – and for leading companies like Centrica, Enel, Royal Dutch Shell, Caterpillar, Boeing, 3M, in a sense that JPMorgan Chase, Bank of America, are rapidly moving to adopt this new generation of technology, so that they candidly don’t become extinct.
Dr Robin Niblett CMG
And when you say they’re adopting this new technology, that they can’t become extinct, the big companies you describe there, were heavily US-focused: Uber, Tesla, Amazon, Airbnb, etc., to what extent do you think there’s something specific, so far, either to the US, maybe to some other parts of the world we’ll get to in a minute, but what is it that’s made the United States such a Petri dish for the emergence of these types of companies? Why is the US, in a way, do you think, some of the companies that are going are American, as well, but why is it going through the extinction and the re-speciation at the speed that it is? What it is about the US that’s done that?
Tom Siebel
I don’t think it’s a US phenomenon. I mean, Royal Dutch Shell is a – is not a US company. Enel, NG, Centrica. So this is happening all over the world, probably in China, at a greater rate than any place else. Now, for whatever reason, a lot of the enabling technologies in AI did come out of the United States and Canada and, you know, I’m not certain why that is, but the – I mean, I just see that as an enabling technology and it is – I think, you know, the large companies in the world are definitely in a position to take advantage of this, and they are.
Dr Robin Niblett CMG
So they should – exactly, so they can avoid the mass-extinction, potentially. Then let me come round to, right now, Conor. This description of the complete change in the way that companies have to think about their futures, companies are made up of people, employees work, it’s something you’ve got to join a company to do. How are you advising, or how have you been advising, and how do you see it, from your perspective? What kind of adaptation are they having to undertake? How successful they’re proving, what do you think are some of the big changes that are taking place, in particular, on the work side and the employment side?
Conor Kehoe
Yeah, well, I’m glad you asked, because some of my colleagues have been looking at this and sor – at a macro level, my former colleagues at McKinsey. So I thought I’d share some of the data that they’ve come up with. I mean, Tom described this trip towards digitisation. They feel that large corporations, on average, are about 20% of the way. There’s 80% to go, and they also feel, by 2030, it could add about 13 trillion to world output. It’s very hard to relate to 13 trillion, and world output’s about 88 this year. But it does mean 1% per annum productivity growth every year ‘til 2030 and that’s a lot. Right now we do about .5% per annum, so just having this phenomenon add one percentage point per annum to productivity growth is quite staggering, and indeed, my colleagues estimate that this will carry on ‘til about 2045. So, as a technology, it is really having quite profound effects.
Now a lot of this is about productivity improvement and their estimate is that about 30% of activity will be automated. But the effect will be spread across many occupations. Only about 5% of jobs will disappear. But in 60% of jobs, at least a third of the activity will be automated. So it’s not as if we’ll say, you know, 30% of this room doesn’t have a job, it’s everybody is affected, to some extent, if that makes sense? And the effect is greatest in routine physical tasks, like operating machinery, fast-food preparation, sort of predictable, repetitive tasks. But also, in data collection and processing and, if I have time, I’ll give you a micro example. This is all very macro, right now. Paralegal work, mortgage remuneration, accounting, back office transaction processing, all of this stuff will be profoundly affected. And less affected will be occupations where the human touch is important, like managing people, childcare and also, physical tasks in unpredictable environments. We’re unlikely to have robot fire tenders, if you like, or indeed, plumbers or gardeners who do – the environment changes, too much of the task changes too much. And then, geographically, the effect is greatest in the developed world, simply because of the economics.
If wages are lower, there’s less incentive to automate, so the – you’ll see it first here and in the United States. And so, if we imagined ourselves in 2030, and we just ask the average Western worker, let’s focus on him or her, what they’d experienced over the last decade. First of all, about 30% of them would either have had to retrain or change job category. So 30% of them would have a story to tell us and in the United States, that would be about 40 million people, telling the story about the profound change that’s happened to them in the last ten years. In the United Kingdom, it’d be about eight million people. And if this retraining had happened within a year or so, we wouldn’t feel too much pain, and they, as individuals, won’t have felt too much pain. But if it stretches out longer, we could have this big dislocation, enough employment around, but not the right skills. So, trying to compress the transition, the retraining, if he or she keeps her job, or the move to another sector, if the job has virtually disappeared, is very important.
And when you listen to these men and women in 2030, you know, the third of the workforce that’s been through this change, some of them will be happy, and they’ll have moved up, you know, they’ll have jobs on the higher end. They’ll have been trained in some form of technology, or something about managing people and frankly, some will be less happy. They’ll have moved from a, sort of, middle income job to a Teacher’s Assistant or an old age care job, and really ,it’s the reasonably well paid middle, which sadly has been the case, I think, for the last 20 years, that’s going to get the squeeze. So, they’ll tell us a tale of either moving up and joy, or moving down and frustration.
Dr Robin Niblett CMG
So there is a risk, in essence, of a further polarisation, and we’ve talked already about polarisation society, we’re living it throughout Europe and the Western world in particular at the moment. So you see a potential greater polarisation emerging here in types of work, I think is what I’m hearing you saying, Conor?
Conor Kehoe
Yeah.
Dr Robin Niblett CMG
I mean, do you have any data, or do you have a sense about how destructive this will be, as well for developing economies? I know McKinsey has a very much a global practice, but there are many parts of the world that we’re expecting, at least, to go through a process, to get themselves somewhere where – who may find that it’s not a case of transitioning from one type of work for another, there is no work to transition into, full stop. Should one be a little catastrophic, in one’s sense of a future and developing world, or do you see actually, as much opportunity, or even greater opportunity there?
Conor Kehoe
In some ways, it’s a bit easier, oddly enough, in the developing world, because there are so many jobs in retail and restaurants that have yet to be created in India, as wage rates go up. So when you start to look at the numbers, of course, technology will make tasks more productive, because there’s no reason, really, for India, for instance, to adopt less-productive technology. I imagine they will, because people are paid less. But there’s so much job creation going on, as it gets wealthier, you just see all these jobs in retail that aren’t there yet, so that tends to swamp the technology effect. The disruption happens in developed countries, where there is already heavy employment, and where you need to switch or retrain people.
Dr Robin Niblett CMG
Where the demography, on the other hand, may provide some help?
Conor Kehoe
Demography will help, in Japan, in Germany, etc., yeah.
Dr Robin Niblett CMG
Well, thanks. Bernice, let me come to you, ‘cause I think maybe we could talk about the opportunity side as well, here. Especially in some of the work that you’ve been doing and your team has been doing in the energy sustainability side of this, and what’s – do you take a relatively – is it an extinction moment, a transformation moment, or how do you see it?
Bernice Lee OBE
Well, we’ll see, I think is the actual two-word answer. But, look, I mean, we recognise the extinction possibility that Tom mentioned, and the polarisation that Conor mentioned and so, we thought why not put different types of extinction and polarisation opportunities together? So we’ve been doing some work around how can we use disruptive technologies to solve problems associated with climate change and energy emissions reductions. So we’ve done it on both the food side, the land use side, as well as the industry side, and Tom very kindly joined us in one of our discussions on the industry side.
You can imagine the technologies are varied, you know, we talk about sensors, we talk about satellite imagery, we talk about machine learning and artificial intelligence, you know, and etc., etc. So we’re not particularly looking at one type of technology, and what I thought would be most useful today is maybe a wee bit mom and apple pie, is to share some of the lessons that we’ve learned from a couple of workshops and research that we’ve done. And I think that that perhaps will help further develop some of the points that our previous panellists have mentioned.
The first thing is pretty obvious, which is that it is often easy for us to have a very lenient assumption about technology, that we find an answer, this will solve a problem. And I think almost extensively ,whether on the food system side, or the industrial transformation side, everyone is saying technology is very important, perhaps more important than ever. But it is still one part of the larger equation, and there are other ecosystem considerations, which we can go back to later. So, how do we make sure that we have a systems level view, when it comes to change and transition, rather than a technology focused one? That’s point number one. Point number two, which I thinks is permeating through all the conversations, all the workshops, all the research that we’ve done, is the need to avoid what I call three kinds of lock-in.
And the first kind of lock-in is, if we actually continue the current economic system, put technology on top of it, are we locking in all the problems that we’ve seen, with the polarisation that we saw, the questions around distribution, redistribution, if we just put technology on top of what we’re doing to date, despite that it might be done by Uber, rather than, let’s say, in the case of the UK Addison Lee, does it really make a huge difference, when it comes to redistribution? And this is not just about artificial intelligence and transport system, it is as true for are we making cell-based cell cellular meats, plant-based meat? Are we still using the same distribution system or are we actually doing it in a way that perhaps introduce more democracy?
The second lock-in that people often worry about is, of course, how do we deal with the governance infrastructure? If the governance infrastructure isn’t designed with inclusion, with distribution of benefits in mind, are there any reason not to expect the monopolisation, weaponization and in certain cases, already we see, in parts of the world, we see both monopolisation and weaponization.
But the last point that, perhaps, for geeks like us, it is extremely important, is the question of intellectual lock-in. The idea that this is actually very engineering driven linear approach to solving a problem and we can learn from all sorts of things, from the Green Revolutions, you know, 50 years ago, which, at the time, lock-in a particular industrial model of investment in agriculture that caused other problems. So unless we are solving problems such as, for example, we are not paying for the cost of social environmental change, for example, unless we are tackling the true cost question, and able our intellectual lock-in to basically lock-down one particular linear approach. Perhaps we can never really genuinely use these new technologies to solve the kind of problems that we have.
So the last point I would make is that everywhere we go, obviously everyone talks about data transparency and how they should be used and how they should be governed. And I think that right now, it strikes us that, you know, all the data richness is determining the level of ambition when it comes to decarbonisation. So companies, with better data, let’s say, to – you know, say a steel plant, if they can only get data for, you know, a small amount, then they can aspire to something like 7% or less emissions reduction, more data tends to help. So there is a dilemma, in some ways, between, you know, the food of the future tends to be data and at the same time, how we govern them, how we use them, is currently, obviously, under debate and, more importantly, how much we would like to use will actually determine perhaps, in an opposite direction, some other prospect that this offers, in terms of public policy.
Dr Robin Niblett CMG
Bernice, thank you, and actually, Tom, that brings me back to one follow-up question to you, and I’m just going to do two or three more and then come out to you, so please do have your questions and thoughts ready. If data richness or data is the food of the future, how we manage data, how we regulate data, becomes absolutely central to being able to take advantage of the opportunities and avoid some of the negatives that are going to come with it. As somebody who’s in the business, what’s your take on these, sort of, different data model – oh, sorry, regulation models that seem to be emerging on data? What has appeared to be, at least so far, a relatively light touch in the US. It may change. Certainly, it’s a big topic in the current Presidential elections. The what’s perceived as over-regulated or European approach, the – I’m not quite sure what you’d call it, the, kind, of the Government will worry about regulation, if it wants to in China, but let’s just focus on getting the most out of the data right now.
Is there a perfect model? Is there a model that you’d like to see that you think would get that social and political balance, alongside the economic opportunity side? Where are you on the regulation debate?
Tom Siebel
Well, I think this is a very important topic. In social media, for example, particularly as it relates to privacy, I don’t think Europe is over-regulated. I think Europe is taking a leadership position here, as it rides to GDPR, I think this is a very good idea. I think, you know, as we think of AI writ large, there will be, you know, enormous social and environmental and economic benefit from these technologies, be it in reducing, you know, greenhouse gas, greenhouse gases or environmental impact, associated with energy production and delivery, precision health, whatever it may be.
At the same time, you know, these organisations that are collecting these data, which are the fuel of AI, okay, that allows us to run these supervised and unsupervised and deep learning models, will be collecting, you know, enormous amounts of personally identifiable information for things like disease prediction. We’ll be able to tell, let’s say, of the population of the UK will – this is within the state of the art today, to be able to predict, or the population of the United States, to identify who is going to be diagnosed with what disease, in the next five years, be it terminal, or not terminal. And the question is, how are our Governments, or our insurance agencies in the United States, going to use these data? Are they going to use these data to restrict access to healthcare? To set premiums? Or, you know, I mean, they will know which of us is going to be diagnosed with a terminal disease in the next three years. Do you want to know that? I mean, I’m not sure I do. I mean, how are these data going to be used? So there’s very, very significant social consequences that we need to think about, some of which are surfacing now in social media. I think is where it’s surfacing first, was the candidly enormously nefarious activity going on by our friends in the social media business, and I think this is a – you know, I am not a big Government guy. I don’t have much use for Government, but I think this is a legitimate role of our Government to regulate, and I don’t know what the answer is, but it does need to be discussed.
Dr Robin Niblett CMG
Well, I’m sure there’ll be some questions on it, so we can bring that – go a little deeper in it, Tom, in a minute.
Bernice Lee OBE
Can I just come in on this point quickly?
Dr Robin Niblett CMG
Yeah, sure.
Bernice Lee OBE
I mean, I just want to say that third party data is always a useful way to keep governments or companies alike in check. I’m thinking, you know, we found out, not because the Japanese Government announced it during the nuclear disaster in, you know, the Daiichi Power Plant, but because the aircraft carrier detected radiation in the open sea and reported back to Chinese, you know, the air pollution was measured by American embassies inside China and therefore, used the data to keep also official data in check. I think there is definitely a role that they could play, provided it is governed properly. So, the question then becomes how do we actually, as Tom said, provide, you know, provide the protection? But I don’t want to lose sight of the importance of the possibility of verifiable third party data, playing a very key role in social environmental affairs.
Dr Robin Niblett CMG
And so, which again, is it how you regulate that verifiable or third party data? One talks about metadata, as opposed to individual data. There are ways of collating it that need not undercut the privacy issue.
I want to just come over, very quickly, to you, Conor. A different topic, just to throw it in there, tax. You know, it’s another big dynamic here, and we’re talking about disruptive, and governments that have been used to accruing income, from very particular types of economic activity, may simply see it not exist and therefore, not in a capacity to be able to use it for those purposes. I mean, do you have any insights on that? I suppose how companies are trying to adjust, or think about it? Or whether they try to av – this is a chance to avoid tax from certain spaces? Or is it an opportunity to be able to get in and think about the social contributions that you make?
Conor Kehoe
Well, I mean, I think there’s two angles to that. I mean, one is not so much about digital itself, it’s about, essentially, Gover –or companies feeling of responsibility towards the environment and the communities in which they live. And most companies now, most large companies, are thinking hard about purpose and how they contribute. At the very least, to attract millennials as both employees and consumers, so that’s an ongoing revolution is too strong a word, perhaps, but evolution that’s underway.
But the other thing I would mention is something that Tom’s book provoked in me. He was talking, in it, about, you know, how sometimes these big changes don’t make their way into GDP numbers, at one point in your book, and it got me thinking, in truth, right now, a lot of the digital economy is a barter economy. So, you get Facebook for free, in exchange for giving them your data. So there are two dimensions, I suppose, to your question. There is tax that’s going to my home country, Ireland, because it happens to be a local for tax environment and they used to demand a 0% tax rate, and now they’ve been forced to demand 12%. So, surprise, surprise, Ireland’s GDP is a lot higher than its GMP, because a lot of American profits are made there. But there is this other aspect of, in a funny way, it’s a very non-commercial world. It sounds an odd thing to say, but people are trading this valuable data for access for free, and it’s a bit like, you know, an old African barter system, in many ways.
Dr Robin Niblett CMG
And therefore, hard for governments to get in and…
Conor Kehoe
And hard for governments to get any hold over it.
Dr Robin Niblett CMG
…be able to – yeah, and it may atomise even more, the more personalised it becomes. You have personalised medicine, but personalised use of your personal data. and again, difficult to know where the Government comes in.
Bernice, you spent quite a bit of time – I know you’re on a couple of foundation boards and so on that have done a lot of work in China, could you just say a word or two about how you think – how you see China thinking about the AI revolution itself? It’s a big topic, obviously, on any issue of technology, but in particular, how they’re seeing the transformation, the extinction risk, but maybe the opportunity for Chinese companies, and others, to really be leaders of the future in these areas.
Bernice Lee OBE
He didn’t tell me he was going to ask me this question. I feel like I need to qualify that. I think that in formal conversations, I don’t feel that there is recognition of the potential job loss that may incur, or the offset of the future jobs that might be created. That’s on the one side. But on the other side, I certainly think that because so much of the economy is yet to be explored, or developed, the LeapFrog has already been seen in, as you all well know, from the payment sector, all the way to food delivery and increasingly, perhaps, in more traditional sector.
But what I’m not seeing, though, in the area that I work on, let’s say, when it comes to steel, cement, you know, industrial revolution, or energy, I don’t feel the same level of enthusiasm for those technology, as I see in the consumer sectors. And I believe that’s partly because those are still very powerful actors, which is the other dynamic, that, you know, incumbents, historically, clearly, have a very strong role to play in many of these industries. I think Tom mentioned a few that have seen their own extinction. But most of them employ lots of people. They have the ears of their governments in which they operate, and it’s usually quite difficult to force them, let’s say, just to substitute technologies, sorry, to get them to substitute technology for jobs, when jobs are cheap as well, as Conor mentioned.
So I feel that the dynamic is mainly in the sector where there’s established interest are less threatened. Whereas, those in the consumer sector, definitely see a lot more, and especially in the fintech space, but again, I mean, I kind of feel that – so recently, I’ve been trying to follow, you know, whether or not Alibaba and Alipay are really expanding around the world and therefore, eating into, you know, the business model of the Mastercards and the Visas, you know, the 2%. And I’m, kind of, observing, like, across Asia, rather than everyone else, sort of, buying Chinese technology, they are saying, “Well, actually, if China can do it, why can’t I? Why don’t I want my own payment system? Why should I get 2%, through Visa and MasterCard, straight to the Americans? And so I, kind of, see a different, kind of, nationalism emerging, based on the use of some of those technology and not in as linear a way as I saw – as I feared, rather, in the first glance. But, having said this, the personal consumer finance sector, which is one of the most vibrant one, using AI, it’s definitely already crashing. So, there’s lots going on, and so, I think that it’s very much an experimental space.
But in terms of Government attitude, I think much more on the consumer side, much less on the established sector. I think slightly in denial about the job question.
Dr Robin Niblett CMG
It’s interesting, and that’s been some of the critique, in a way, of some of the best investments so far around AI, at least is we see on the consumer side, has been in the consumer retail space quicker to make the profits. Some of the tougher calls are in these big transformational industries. Yeah, come on in, and then I’m going here, yeah.
Conor Kehoe
A very quick comment on that, I mean, because what you’re saying about China is also true. If you look at the industries that are ahead, I classify them slightly differently. I’d say there are industries that deal in immaterial things, there’s services industries, because it’s just so much easier for them to, essentially, translate marketing and administrative functions to digitise them. And the industries that are dealing in manufactured products, so big swathes of manufacturing, and indeed, the pharmaceutical industry, which is quite a modern industry, are quite far behind dealing, as they do, in physical products.
Dr Robin Niblett CMG
Tom, do you want to say something? Because actually your company, I think, is trying to provide this technology to those big material producers, as I understand it, energy companies and others. Are they behind? Are they – do they need to, sort of, bring in companies like yours to help them transform, as opposed to them being driven by transformation themselves? What’s going on, on the material side?
Tom Siebel
Well, again, I mean, if we look at the information technology industry globally, in 1980 it was a 50 billion dollar business. Today, it’s three and a half trillion. In five years, it’ll be eight and a half trillion. It’s growing at an accelerated rate and this acceleration is all about Elastic Cloud computing, AI and IOT. So there’s absolutely no question about – this is not an if, okay? I mean, this software segment alone will be a quarter of a trillion dollar software segment in 2023. So this is the fastest growing software market in the history of the software industry.
Now, the people that we’re seeing adopt, I mean, Royal Dutch Shell, globally, Rosneft in the UK, globally, Saudi Aramco, globally, and now, in Rome, NC in Paris, Centrica in the UK, Bank of America in the United States, JPMorgan Chase in the United States, CAT, 3M. So we see it across – Ericson in telecommunications, so this is certainly not specific in the United States, it’s – it is not – I believe we’re seeing it in services, we’re – I know it, okay? We’re seeing it in manufacturing. We’re seeing it in oil and gas, and we don’t – we’ve elected not to do business in China, but if you take the time to read the 15th five-year plan on Wikipedia, I mean, the roadmap is very clear. Client training is absolutely the largest adaptor of these technologies across all industry segments.
Dr Robin Niblett CMG
Right, let’s get some points in. As we’ve got a panel, I’m going to take two or three, and then we’ll take it back in conversation. Who wants to go first? Who’s got a point or a question? Starting here, introduce yourself, please, and I’ll get some others in. Hopefully, you won’t be too silent an audience, yeah.
Noel Hadjimichael
Thank you very much. Noel Hadjimichael, Member of Chatham House. If data is the diet or the food of the present and the future, given the distribution, and given that it’s very likely to be a very uneven distribution, will free pluralist societies that we know, and value, find it much harder to manage this digital transformation than less free and more authoritarian societies that place less value on consent?
Dr Robin Niblett CMG
I’m glad we got to that question, ‘cause we talked a little bit about that beforehand that that might be something that would come up here. Other questions? Other points? Otherwise we can go round that way. Yeah, please, gentleman here. The microphone’s coming.
Arvin
It’s on a different topic.
Dr Robin Niblett CMG
Sorry?
Arvin
It’s on a different topic.
Dr Robin Niblett CMG
That’s fine, no, a different topic’s good, ‘cause it’ll also inspire other people with different topics.
Arvin
Arvin, individual Member. A question for Conor, actually. You were talking about certain classes of jobs that are at risk. For example, those in contact with people, and you mentioned management as a role as not being at risk. Would I be right in arguing the opposite, actually, that obviously, childcare, care for the elderly, that’s a different story. But management, a machine Manager would be much better. It would be unbiased. It would take the plethora of data that comes to it, analyse it much quicker, make an unbiased decision where the investment should be, where the lack of investment should be. Don’t you think that management could be, actually, the next set of roles at risk? Thank you.
Dr Robin Niblett CMG
We’re all going to become Artists very quickly, I think.
Tom Siebel
Why don’t we make this, Robin, at the risk of – yeah, why don’t Conor, why don’t you – let’s do it and let’s make this more interactive.
Dr Robin Niblett CMG
Yeah.
Conor Kehoe
It is very clear that, you know, algorithmic decision-making, where it can substitute for human decision-making, if you can get an expert to write the algorithm, the algorithm’s better than the expert, because we, as human beings, have good days and bad days, and that’s pretty well known and you’ll see that more and more. But when it comes to managing and motivating people, as against algorithmic decision-making, as yet, and in any event, the technology isn’t up to it, people still do that better. So you make me make quite an important distinction, so I think there’s still plenty to go for in managing and motivating people. But, as I say, if you can turn it into an algorithm, the decision-making, it’ll be a better decision-making machine than human beings.
Dr Robin Niblett CMG
That’s your point about combing things as well and combining skillsets, you might have, well, I suppose access to the machine learning information, guiding you on what to do with people. You then apply your human judgement, for the two or three people, where you want a different hand, yeah. So, as we’ve come back here, sir, let me – we have that very big topic on the table about whether plural societies and polities are we going to find it easier or harder to make this adjustment? Where do you come down on that, Tom?
Tom Siebel
Well, you know, in China, we have a very, you know, top down command and control totalitarian society and I don’t think there’s any question, they will adopt these digital technologies at a greater rate and they are and I think the social and human costs are going to be, you know, pretty onerous. You know, as you – like, I mean, you’re well aware, you know, the social compliance scores that they have going on in China now, this is pretty scary stuff, okay? You know, what they’re doing, I mean, there is a – they – when we get into smart energy and the internet of things, I mean, China is adopting these technologies at a very, very rapid rate and this, you know, NRDC has a way of writing these 12th, 13th, and 14th five-year plans that just kind of happen. So, it’s – I think it’s happening at a more rapid rate there, than any place in the world.
Dr Robin Niblett CMG
So, you wouldn’t buy into the theory, then, the US, with its more chaotic political system and the more competitive element, more bottom up and less top down, which has done very well on technologies, looking at your career, and other people’s careers, that the Chinese system is going to start to outplay it, outthink it over time? Are you down on America or are you just up on China as well?
Tom Siebel
Kai-Fu Lee argues that in his book called AI Superpower, very eloquently, okay? But AI Superpowers, but I think that, you know, I mean, we – as it relates to AI, I mean, we are at war, okay? And it’s full game on, and it’s right now. I mean, the Chinese are penetrating the European Grid every day, the US Grid every day. They’re implanting viruses and they have the ability to turn the Grid off or turn the financial system off. They have robots, they have robot satellites. There’s a book called Shadow War, really interesting book, about the satellites they have, these killer satellites that can just take the GPS system out. You take the GPS system out, there are 24 satellites, GPS satellites out there, okay? And the European and the US Military are blind, can’t navigate, can’t communicate, and can’t pull the trigger, okay? And so, we have the Chinese today in top down, very organised way, are spending 20 billion a year, going to 60 billion a year, in advancing the tools of war, as it relates to AI, and I think what we have is a – is the ultimate test of two political philosophies and particularly in the United States, we have this very messy free market capitalist society where, you know, this work is done in garages in Palo Alto, and storefronts in the Bronx. And so, even Vladimir Putin said, in 2017 in Sochi, he said, you know, “Whoever wins the war on AI, dominates the world,” and I believe this is true. And I think we’re seeing the test of two fundamental political philosophies and let’s hope we don’t lose, because it’s not going to go well if we do. But I don’t, for a minute, you know…
Dr Robin Niblett CMG
Yeah, so let’s go that way.
Tom Siebel
You know, I don’t, for a minute, think that, you know, that, you know, the innovation of Western Europe and the United States is – I mean, they have a way of creating things out of nothing and, okay? And winning in the long run and so, let’s keep our fingers crossed on that.
Dr Robin Niblett CMG
Okay, we’ll keep working on it. I mean, you said on a – we’re at war, you might say, ‘cause we’re at war and AI is part of it, I mean, it might be a cold war but, you know, rather than this, kind of, that we’re at.
Tom Siebel
And the Grid every day, I mean, they took, you know, 21 records of – you know, they did the entire personnel system from the Office of Personnel Management in the United States for 21 million people, including everybody who’s ever applied for and been granted a security clearance. And they can’t create the Grid every day, if this is not war, what is it?
Dr Robin Niblett CMG
You wouldn’t think the US are doing the same thing and we don’t know about it?
Tom Siebel
No doubt.
Dr Robin Niblett CMG
I would – I’d be highly surprised if the US wasn’t doing exactly the same and, if the UK could, it would be trying to as well, or maybe helping the US.
Tom Siebel
One is a documented fact and the other is speculation.
Dr Robin Niblett CMG
Okay, and so…
Tom Siebel
Well, one is a well-documented fact, the other is speculate, okay?
Bernice Lee OBE
But I mean, I think that there is no question there are reason – I mean, I think Lee Kai-Fu’s book was excellent. But one of the reasons I think that is not emphasised enough is the sheer volume of people in China and therefore, data source, but of course enable a much larger faster growth in the consumer market that has come online.
But actually, I was thinking – but your question is about whether pluralistic societies are more or less like a Manager transition. So, I – you reminded me of a story or an argument that I think Larry Summers makes quite a lot, and I heard it last time at a China Development Forum, with Robin in fact, a couple of years ago, when he talked about how, when the US transitioned from the previous more agrarian economy into the modern, you know, production-driven fort-like economy, it was facilitated in no small part by the Great Wars and therefore, the Great War, the Second World War, and therefore, the GI Bill that put a lot of farm boys, at the time, from farms into universities, through subsidised education. So subsidise – and many of them became Teachers, Lawyers, all sorts of things.
So, even in a very pluralistic economy, that managed the last transition and won the last transition very well, there was a huge amount of Government intervention, facilitated, obviously, by a World War at the time that actually made that possible. So, I guess I – whether – I’m not entirely sure that it’s a question of what system, but it’s a question of how you use a system to create the opportunities for people. It strikes me as being as important as what the starting points would be and no question that one leads to one outcome and the other perhaps less likely to that outcome. Though, I’m not entirely sure, if you know what I mean, we now know what system leads to what outcome anymore.
Having said this, though, I do think that when it comes to, therefore, governance of knowledge and data, that would be the, you know, in the civil societies then, civic society, as opposed to military society sense, that would be the challenge. You mentioned drugs and medical diagnostics and I was fascinated by the story from Memorial Sloan Kettering, you know, when they started a start-up recently and put 20,000 slides that were diagnostics of Doctors, who made, frankly, probably quite a good living for 30 years, on the back of that knowledge. But now, this is all going to be put in an algorithm and learning will happen and we’ll learn more than these Doctors would. And then there’s the patient who, whether they sacrificed their life or not, who’s data or that, you know, whatever components that we end up analysing comes from, just to, sort of, figure out what that really means, in terms of distributions of benefits and how we manage that is mindboggling, let alone something much greater than the question that you asked, which is, you know, how would societal system interface on top of all these challenges?
So, I would say, if we are mindful, and do what is needed, we have a better chance, but I’m not entirely sure that it is about, you know, free or not free, and freedom of choices, when it comes to, you know, governance.
Dr Robin Niblett CMG
As with seen with authoritarian states, that mostly have not taken advantage of this revolution, unlike China which has. So you’ve got plenty of authoritarians out there that haven’t managed states that have not managed it. Yeah, sorry, there was a question here, yeah, please and anyone else, yeah. The microphone’s right there.
John Cooke
Thank you, John Cooke from TheCityUK and a Member of Chatham House. I have a question about regulation. I was thinking particularly of financial services regulation, but it applies more generally, I think. Conor mentioned the changes in role within organisations, and I can see that on the one hand, if one’s looking at regulation of financial services, there are certain key ratios, which are likely to remain constant, and which will be important for testing the soundness of businesses.
On the other hand, if all these roles are changing, then who should the Regulator look at as a fit and proper person in a particular area of a business? That, I can see, is going to be a, you know, there’s going to have to be a new understanding of where critical decision-taking is really happening. And I’m not sure whether the same applies to economic regulation, with anti-trust and so on, and we’ve spoken about text and how you measure where activity is happening, but I feel there’s a collection of issues round that on which I’d be interested in your comments.
Dr Robin Niblett CMG
Okay, so on regulation, I mean, do you want to take that one on right now, and then, yeah?
Conor Kehoe
My immediate reaction is, when it comes to the, sort of, high level regulation of ratios, some economic ratios, since they tend to be vested in the very senior management, that won’t change too dramatically. However, the behavioural regulation, which is what has upset the financial system recently with mis-selling etc., that actually is a real issue. I don’t have time to go through it, but we – what’s happening with many financial products is the decision to grant them, and let’s say a mortgage, is being automated, and is driven by AI, so the question then becomes is, what kind of behaviour does that lead to? And in fact, do the AI algorithms lead to behaviour that you mightn’t find desirable, because they look to the past, they see who’s been a successful mortgage grantee in the past, and they apply that to the future. So new populations or disadvantaged populations find them – that the systems itself perpetuating their disadvantage. So, interestingly enough, I think that’s where the Regulator will have to do something.
Dr Robin Niblett CMG
Let me come in, a question here at the front.
Esther Naylor
Esther, Chatham House. You’ve talked a lot about automation and disruptive technologies. In terms of deep learning and unsupervised learning, which sectors do you think that they have the most potential of application?
Dr Robin Niblett CMG
Where will the deep learning have the biggest impact you were saying? Which sectors? Well, Tom, I’m going to have to start with you on that one, I think.
Tom Siebel
Okay so I’ll comment on deep learning. So deep learning is this area that we call neural networks and given – and we measure the accuracy of AI models by – on two axes, and they are precision and recall, okay? And the – and so we can put a number on how accurate a machine learning model is. And, given enough data, in every case, a deep learning model, a neural network will provide greater levels of precision and recall.
The problem with deep learning is, it is a black box, okay? So it’s inexplicable and so, for many, if not, most applications, it’s simply unusable, okay? You know, for example, for credit applications, and we do the automation of credit applications at Santander Bank, we do it at Bank of America. Now this is – it is not a machine making the decision, okay? Okay, what it is, is it is a machine just aggregating the data, okay? And providing it to the decision-maker to make the decision and it just ex – it compresses the time to aggregate the data from, say, 70 days to seven, so your loan gets approved quickly, but it’s the same information, provided to the same decision-maker, who makes the decision.
Now, the trouble is, with using traditional machine learning, what’s called supervised learning, or non-supervised learning, we can provide the decision-maker with an evidence package, so it can see exactly, you know, why this – we’re giving you this credit score. Using deep learning, it’s inexplicable. So for most applications that we deal with, in the commercial industrial sector, you – candidly, you know, deep learning doesn’t work. If we need to, like – if we’re talking about – I mean, this is another project that we’re working on, for example, the Space Command for the United States where they’re you know, identifying objects in outer space or doing target acquisition, okay? Is it a MiG, or is it a 737? These are applications, for which deep learning actually will work, and people are willing to accept it, because the precision is just so high, you know, ten times out of ten it’ll distinguish between a 737 and a MiG, and it’ll get it right.
Conor Kehoe
Just on that, I’ve tried to program some of these things and I tell you, it’s mystifying because, let’s say you’re trying to get it to recognise this is an automatable or a car in a picture. If you go a few layers down in the deep learning algorithm, there are these wonderful abstract pictures, which have some of the essence of car-ness in them, that’s what you see, and you just can’t relate to it at all. And, you know, as you say, you can’t explain what’s going on, but the output’s very accurate.
Dr Robin Niblett CMG
Right, got a hand up here at the front here, and one at the front here.
Bernice Lee OBE
Another one behind that one.
Dr Robin Niblett CMG
Another one behind this one. Okay, yeah, [inaudible – 50:33] first, yeah.
Member
I’m [inaudible – 50:34], a Member and Behavioural Economist, and my question is, in some cases, humans are much better decision-makers and, in some cases, computers are much better, what is that line? Do you think that, like, this decision must be made by human, for example, like autopilots in the plane, what if something goes wrong, like the Hudson River landing, computer couldn’t make it. Human was much better landing on the sea – river. What’s that level of human versus machine?
Dr Robin Niblett CMG
And will it change, I would have thought, as well? It must be changing all the time, with the difference of responsibilities. Tom, do you have a view on that?
Tom Siebel
Well, I think this gets to autonomous vehicles, okay? And there are a lot people, not me, okay? Who believe that autonomous vehicles are going to happen tomorrow, okay? And the trouble is, and particularly when we get into computerated vision, when we get into these issues of facial recognition, this is very scary stuff, okay? This facial – these facial recognition algorithms are very easy to trick. MITRE Corporation has algorithm where if you put on a – you put a button – lapel button, and they’ve developed this, they haven’t published it yet. MITRE Corporation have a Research Institute in Washington DC and if you wear this button, every facial recognition algorithm on the planet will identify you as the CEO of MITRE, okay?
There are stickers that if you put on stop signs, okay? That if you put a sticker on the stop sign, okay? That every autonomous vehicle will not see the stop sign. So, you get in – so, I personally don’t think – I think we can get a computer to play Go better than a human, play chess better than a human, maybe even bake a cake better than a human, but you can’t get a computer that’s going to play Go, play chess and bake a cake, that’s not going to happen soon.
Autonomous driving, I mean, this is a tough issue, I mean, what decision is the computer going to make? Somebody is going to die. The driver’s going to die, or the old man with the cane is going to die, or the woman with the stroller is going to die. Somebody is going to die, okay? What decision is that computer going to make? And whatever decision it makes, it’s going to be wrong, and then we’re going to have hearings in Parliament and it’s going to be the end of the story.
Dr Robin Niblett CMG
And I suppose even if one is wrong, and normally 100 people die in car accidents, the fact that the 100 are done by humans, but the one is done by a machine means the one number the machine is the one that you focus on.
Tom Siebel
Whatever decision it makes, is wrong.
Dr Robin Niblett CMG
Yeah, Bernice if you want to come in now, and then I’ll come and…?
Bernice Lee OBE
Yeah, I do, and I think that I completely agree with what Tom just said, but I also think that it, kind of, depends on what do you mean by driving, and what do you mean by transportation? And, you know, look, I mean, I’ve been joking with my former colleague, Phyllis Preston, for a long time about how, you know, in London our average driving speed is eight to 12 miles. In almost any circumstance, any testing, you know, eight to 12 miles, pretty safe. So we just need to, really, get rid of all human driven cars and get rid of the idea that we should speed – able to speed up, ‘cause I’m, like, why do we need to accelerate? You’re not trying to rob a bank, right? So, we, kind of, get rid of all the prec – you know, our preconception about what driving in inner cities should be, we could do it relatively safely and you’re completely right, I’m not going to be able – no-one will be able to teach a computer to kill, you know, even to – I mean, I thought the only command it should be allowed is to say, “Kill me, not anyone else,” and that I felt like that’s the decision that a car should be empowered to make, when you drive, you know, when you put yourself in an alternat – autonomous vehicle. But I, kind of, feel like you’re right, this is why the technology in itself may not be a solution, unless we think bigger around how do we organise transportation? And I can imagine there are areas where you can have autonomous shared vehicle, but it’s just not in the way we understand driving and cars today.
Dr Robin Niblett CMG
This is your point about not imposing AI on top of existing ways and a structured way of thinking about it. This – now, you see, what happens, all the hands start going up at the end. So, three down here. We’ll run five minutes over, if everyone’s alright with that? Because we have a – we started a little late, yeah?
Marjorie Buchser
Marjorie, Chatham House. So we talked a lot about penetration of those technology in private sector and industrial processes, but I’d like to hear your thought about penetration in Government or not-for-profit. So, of course, Government can regulate the sector, but also, integrate it into its practices and so far you have smart cities, but it’s not really widespread. So my question is, how do you make smart policymakers that actually use our technology, as well as the private sector?
Tom Siebel
How do you make smart policymakers? That, I think, is like – that one I’ve got to leave to Conor, I’m not going to touch that.
Conor Kehoe
Happy to settle for official administrations.
Tom Siebel
Absolute morons, sorry.
Dr Robin Niblett CMG
But if no-one else is coming on that point, and penetration of AI into Government, though, doesn’t just have to be decision-making, but obviously, it’ll be into healthcare, it’ll be probably into delivery of Government services, yes? And all of that space. Do you – but…?
Tom Siebel
Defence.
Dr Robin Niblett CMG
Defence.
Tom Siebel
Intelligence.
Dr Robin Niblett CMG
Well, yeah, where does it go first? And I suppose, well, it goes where the money is, initially, and where there’s sense of the need and the drive.
Tom Siebel
The first place it’s going is absolutely defence, okay? And for I think for a very good reason, but we get into, you know, I’m not familiar with what’s going on in Europe. I’m pretty familiar what’s going on in the United States. I have some familiarity what’s going on in China, and defence is absolutely the first application of these technologies.
Dr Robin Niblett CMG
Where Government, exactly, is getting involved.
Tom Siebel
But it will be in healthcare, I mean, think about tax. Identifying tax fraud, okay? Healthcare fraud, I mean, this is a national application of AI where we could very easily, I mean, I don’t know what the levels of fraud are in the UK. I assume they’re similar to that which they are in the United States, as it was to health and tax, and it’s huge and so, I mean, we’re all paying for that and that’s a national application, AI.
Conor Kehoe
And I think, actually, the place I’m a bit disappointed about the progress, and because the tax authorities are doing a reasonably good job here with it, is in the healthcare system. Because we have this huge uniform healthcare system, and we’re always dreaming about how, if we could get all the files online and, we could just do wonderful work comparing, you know, the background and demographics of certain diseases and move to prevention or anticipation of diseases. But it has proven very difficult to get the professional body to submit these data together, you know.
Bernice Lee OBE
And on health, actually, can I pick up? I went to a presentation by a AI company on health and it’s Babylon, I should say, because they are providing universal health service for the Government of Rwanda. And, you know, I think in the first week, something like – I can’t – I’m making the numbers up now, 200,000 people sign up or something and they promised to service the whole country.
Now, we also have them in the UK. Babylon Health is a AI – I don’t know what would you call it? A sort of substitute, and what – and so, I saw a presentation of the present – of the machine. So, in – they said that the AI Doctor, in the third instance, already scored 96 points in the test that a normal Physician will take. The Royal College of Physician when, I think, an average, for humans, and I can’t remember exactly the number, but let’s say in the 80s. And they consistently did so and the machine is not meant to replace healthcare, but to take away the bit that is not needed. And it’s also what is interesting to me is not, like, I imagine here, in Rwanda where you have, you know, everyone has a phone because obviously, not everyone has a smartphone, it’s about putting a really good computer in the hands of a Community Nurse, who could sniff out the cases and then set appointments for Doctors, when these cases are serious. So I think, again, it’s a bit like, depends on what you think of healthcare and if it is about healthcare, I can certainly see that there’s a huge cost reduction for everyone involved and that perhaps is going to be welcome.
Conor Kehoe
Yeah, there’s two things, if I might say, going on there. One is, you’ve seen that with Pilots, once you introduce process, this is back to behavioural economics, you already get great benefits, because even if that Doctor knows quite a lot, if he or she doesn’t follow a clear process, the error rates go up. So process gets you a lot of the way and then, of course, on top of that, there is an ability for AI, for instance, to recognise and you can now do it online, if you’ve got a little blemish on your skin, you can check out whether it’s cancerous or not, so that’s a wonderful application, you do it yourself tomorrow. So there are two things going on, one is just having better processes to improve the psychology of decision-making, and then there is the technology to help on top of that.
Bernice Lee OBE
And you’ve got to spend time writing up the reports, so the machine has it.
Conor Kehoe
Yeah.
Dr Robin Niblett CMG
Right down to the business of people wrong – writing wrong prescriptions on paper and peo – the amount of people of die every year from that. So, I mean, there must be so many gains that can be made. But, in the end, the return element, what you said earlier about the returns that are being provided for the intangibles the, sort of, bartered goods, that’s where the IPO returns seem to be existing heavily right now, and often not in the social sector.
I said we’d finish at five past, but we have two people waiting here. So you just, both your questions, very quickly, and I’ll give a chance for everyone to wrap up and anything they didn’t get to say, they can say.
Member
Thank you. On the Russian Government tax, I’ve got one question. Because AI’s been long used for analysing data, but now it’s very new, you can use it to generate data, fake data, deep fake, these sort of things, and I’m wondering, are we going to a world where data stops being as valuable as it used to be and what sort of solution would there be for this?
Dr Robin Niblett CMG
Wow, I love it. You know, AI depends on the data, but then, may be you have fake data generated, which makes it less valuable. I mean, that’s an interesting frame turner to the end. Yeah, last comment.
Jan
Jan, Member of Chatham House. So I was wondering as US you said leading the AI scene, how it can be also raised aside of accountability, in terms of, let’s say, the technologies being used by the Governments, which is not democratic and free. But, I mean, I’m just saying, like, as like Google Maps and they have all our data and when they have all our data, it can be chased and known by those governments and be used against us. So – but shouldn’t Google still be kept accountable, from US, even though we’re in Africa somewhere about how it – how our data is going to be used?
Dr Robin Niblett CMG
Thank you. I mean, the issue we didn’t get into was ethics around AI, the accountability of all that data and how it’s used and how it’s supervised obviously is part of that. Closing set of comments. Let me go in reverse, Bernice, anything you want to pick up on the fake data and on the accountability? Just a quick comment from each of you.
Bernice Lee OBE
Yeah, on the last point, perhaps, I think that bad people will always find their way to get us. So, in a way, I’m not willing to therefore say that I should be less transparent for that reason. I’d like to think that I – we would do other things right. So, you know, if anyone wants to steal and, you know, you can already track me down between my mobile phones, my Oyster card and my credit card, probably, easily. You could do so ten years ago. So I, kind of, feel like we need to harness the good parts and be very clear about where the no-go areas really are. Because, for – you know, pretending that bad things will happen may end up legitimising the small bad things happening.
Dr Robin Niblett CMG
Okay, so that’s closure. Yeah, go on.
Conor Kehoe
Data I hadn’t thought about much, but I’ve been thinking about Tom’s bartering point that he made in his book, and maybe ultimately, good data will have a commercial price and there’ll be a big pool of – and so it might start to stratify. And then, on accountability, I mean, I tend to agree, I’m glad the EU has come up with GDPR, and it may not be perfect, but it’s a start and the EU is about the same size economy-wise as the US, so it’s a big player in the world economy. It’ll improve, I hope, over time. Somebody else will come up with something better, but it’s great to see a start happening there.
Dr Robin Niblett CMG
Tom, any last thought on these points about bad data, good data?
Tom Siebel
I have nothing to say, and I just wanted to – I just want to say thank you, everybody, for the courtesy they extended us and it’s a great privilege for me, and I think for all of us, to be here with you this afternoon, so…
Dr Robin Niblett CMG
Successful CEO does his own wrap-up. So I won’t even to try to keep up with that. I’ll simply say what Tom said. So give a strong round of applause to everyone [applause].
