AI is Disrupting Canada’s Labour Landscape

Speaker 1 [00:00:01] Hi, it’s John here. You have to be in the middle of some kind of epic digital detox to not know that this is the year of AI, especially AI adoption. Not a week goes by without one of the platforms unveiling a new AI tool that’s transforming how we live, think, share, and even play in the digital universe. The Canadian government stepped into the fray in its recent budget, announcing $2.4 billion of new funding to support Canadian computing power and help Canadians take advantage of this new age. The goal is clear to spark innovation and boost productivity, and all that money actually may not be enough. I got to spend a weekend in early March, just outside of Silicon Valley, with some of North America’s leading AI thinkers and doers, corporate leaders, investors, techies and regulators who are trying to figure out where all this is going, and also think ahead to a possible convergence of humans and machines that will change, well, pretty much everything. It’s a bit out there, maybe even sci fi ish, but there are plenty of questions about the world of today and what AI is doing to our lives and our jobs. One of the most interesting voices at the retreat is my guest today, David Autor, a renowned labour force economist at MIT who spent his career studying the impact of technology on jobs and the economy. We all may have different views and even emotions on the consequences of AI, ranging from fear to excitement. So spoiler alert, David thinks we’ll be fine if we prepare. So how do we prepare ourselves and the workforce for what’s to come? What does AI mean for white collar work in the middle class? And how do we ensure AI is constructive and not destructive? We’ll have a very human conversation on those questions and much more. This is Disruptors, an RBC podcast. I’m John Stackhouse. If you haven’t heard of David Autor, you should. He’s a brilliant economist, public policy scholar and professor at the Massachusetts Institute of Technology in Cambridge, just outside of Boston. He wrote the book The Work of the Future Building Better Jobs in an Age of Intelligent Machines. He’s released groundbreaking research on Automation’s impact on jobs, the nuances of skill-bias technological change, and the pivotal role of education and policy in shaping the workforce of tomorrow. David, welcome to disruptors.

Speaker 2 [00:02:40] Thanks so much for inviting me on the show.

Speaker 1 [00:02:42] We’re talking about AI, and it’s hard not to start with some words from Elon. So I’m going to quote an interview he had with the British Prime minister not so long ago. And he referred to artificial intelligence as the most disruptive force in history. Even went on to say that there will come a point when no job is needed. True or false?

Speaker 2 [00:03:04] False. We’re going to have lots of jobs. We shouldn’t actually be worried about the quantity of jobs. We should be worried about the quality of jobs, what type of expertise they use, how well they pay, whether they provide economic security. And that’s not simply a function of AI itself, but the way we use it. This is a technology. It’s not a living entity. It doesn’t get to decide how it’s used. And we have a lot of agency about whether we use it mostly to replace workers and automate things, or whether we use it to enhance the value of human expertise and create capabilities that we don’t already have.

Speaker 1 [00:03:38] David, you’re a great student of history, and I’ve been following you for years. In fact, you’ve helped open my eyes to the fact that over a couple of centuries, in fact, technology has never led to less jobs. It’s usually led to more jobs and often better jobs. I work at a bank which our sector invented, the automated teller machine that was supposed to eliminate thousands and thousands of jobs, which of course it did in one respect. But the sector employs more people and arguably with better jobs, better paying jobs, more productive jobs than ever before.

Speaker 2 [00:04:09] That’s not to say that it’s all a win, right? These things are quite disruptive. And the industrial revolution, the Luddites who rose up against the power frames, they weren’t wrong. Technology wiped out the value of their skills, wiped out their livelihoods. And the last 40 years of computerization have also been actually very hard on middle class jobs that displaced a lot of workers from office, clerical, production and administrative support. And I think I will also disrupt. So it’s not that there’s nothing to be worried about. I think people are focused on the wrong thing. If we look around all rich countries right now, they’re short of workers. Not short of jobs. That’s not the question. The question is, what type of work will we be doing?

Speaker 1 [00:04:50] So to the people whose jobs may be disrupted, even eliminated, is that just the price of technological progress, or are there better ways at transitioning and transforming work?

Speaker 2 [00:05:03] There are better ways. And different countries do this much better or worse. For example, during the China trade shock of the 2000, the U.S. lost more than a million manufacturing jobs to China, and people were really damaged by that and communities took a very long time to recover. But in other countries, similar declines in manufacturing don’t have as catastrophic consequences, including in Canada, by the way. And in Denmark, which spends about 3% of GDP on worker training and activation, which is an order of magnitude more than the U.S. spends, there’s very little scarring. So some disruption is almost inevitable. But how we support workers to make those transitions, what new opportunities open? And also the speed at which that change occurs, those things are influenced very strongly by policy and by investment.

Speaker 1 [00:05:54] The Canadian government has just put down more than $2 billion for an enhanced AI strategy, a significant amount of money, especially in tighter fiscal times. A lot of that is going to the tech side of things to compute and supporting research. What do we need to be mindful of to ensure that $2 billion isn’t eliminating jobs, but actually enhancing?

Speaker 2 [00:06:18] I think it’s a question of how that money is directed. If I were in charge of this stuff, and obviously I’m not in charge in either the U.S. or in Canada, I would want to take some moonshots for that. The health care sector is an area that could benefit enormously from AI, not just in the quality of care, but also the type of work. And there’s always an unlimited demand for health care. There’s long waiting lines. It’s expensive. And we could improve the opportunities for workers in this sector, meet consumer demand and there would be no adverse job consequences. And of course, that’s public sector money. So all the more reason to use it well. It helps have a vision of what you’re trying to accomplish. And I do think the Silicon Valley vision and you and I both know this directly, having been at conferences together recently where we saw this firsthand is really about eliminating labor. They keep saying we’re going to eliminate scarcity. Of course they don’t mean all scarcity, right? So all the AI in the world is not going to create more beachfront property in Monterey, California. It’s just going to reduce labour scarcity. But reducing labor scarcity really shouldn’t be our goal. Labour scarcity is actually an achievement. In fact, in most rich countries, labour is paid somewhere between 55 and 65% of all national income. So we should be trying to use labor better, not trying to undermine it. That’s not the type of scarcity we need to get rid of. Because if you said we got rid of all labour scarcity, saying we’d be getting rid of all jobs, I don’t think that’s feasible. But I also don’t think it’s desirable. It’s not the goal we should be shooting for. I think many people think it’s very natural that the way you use the technology is to automate stuff, right? You have a new technology. What can we do that we’re doing by hand that we can now do with a machine? But automation is actually a very small part of what we do with new technology, and not where most of the benefits come from. If we were to go back to ancient Greece and automate everything they did 3000 years ago, that wouldn’t give us modern Ottawa. That would give us ancient Greece without horses, right? Most of the most valuable uses of technology are because it instantiates new capabilities. It simply was infeasible. Without those tools, our lives are utterly transformed by technology, but not so much because of automation, but because of all the capabilities that we produce with it. And AI has certainly lots of that potential that can change the way we research. It can change the set of skills or tasks that people can do. It’s a really great decision support tool. And so that’s what we ought to be using it for. Rather than just thinking of what can we eliminate.

Speaker 1 [00:08:50] I think I’m quoting you here, David, when you said the microscope didn’t eliminate any jobs, like, think of all that it led to.

Speaker 2 [00:08:57] Yeah. Even more specifically, the scanning electron microscope didn’t eliminate the job of looking at subatomic particles. We just couldn’t see them before we had the scanning electron microscope. And yeah, it’s true for flight. Flight didn’t eliminate the way we used to flap our arms and go from place to place. We just didn’t fly before we had planes.

Speaker 1 [00:09:17] There are many AI proponents who say generative AI particularly is different. It will allow machines to replace human activities, including whole job categories. Do you see it differently than previous transformative technologies?

Speaker 2 [00:09:34] Well, first of all, generative AI is different from other technologies, and in fact, all technologies are different from one another. Generative AI certainly can do things that previous technologies couldn’t do. Computerization prior to was extremely good at carrying out formal rules and procedures in a way that used to require a lot of education. It was very good for people who are decision makers and professionals who have. Having access to information and analysis is extremely beneficial. It was obviously not particularly helpful to people whose primary job was, you know, typing and filing and copying or doing repetitive assembly line procedures. So generative AI will certainly complement and substitute for a different set of activities. What it’s really good for is weaving together information, rules and kind of judgment experience to support decision making, right? Whether it’s reading x rays, doing some types of writing, or even a skilled repair, having a kind of a copilot in many things you do could be quite valuable. So it certainly will eliminate some work. I don’t mean to suggest otherwise. There’s no question that certain types of decision-making tasks of managerial tasks, judgmental tasks, will be automated, right? So it’s reasonable. Amazon no longer uses people to do inventory projection. That used to be kind of a task where you combined information plus judgment and sort of intuition to make those decisions. The machines are really good at that. That’s true. On the other hand, I’d like to think, and I think this is true, that we can use these same tools to enable people to do some of that hard work. So if you think about what is expert work, expert work is using judgment to face high stakes, one off problems. So your figuring out how to treat this cancer patient, or you’re architecting a piece of software, or you are rewiring a house, or you’re landing a plane under challenging conditions, right? All these things, you have some formal knowledge. Obviously you need that. Then you have lots of experience and you have to apply. You have to translate from that formal knowledge and your experience to this one case to make the right decision. And there’s very high upside and very low downside or a high cost to making the wrong decision. If you a skilled trades person, that’s what you’re paid a lot for. If you’re paid a lot as a roofer, for example, it’s not because you’re good at pounding nails, right? It’s because you know how to put on a roof that doesn’t leak. AI is very good at supporting expert work. It doesn’t necessarily always make the right decisions, but it can help guide decisions. So, for example, a medical setting, it could say. This set of symptoms would be consistent with the following. Had you consider these diagnoses and could also say, hey, don’t prescribe those two drugs together, they have a negative interaction. That doesn’t mean you don’t want a nurse practitioner or a doctor doing that work, but having them guide it in this way could enable more people to do that work.

Speaker 1 [00:12:21] Some would suggest that that supports the idea that this will create mass expertise. And if there is mass expertise, maybe there’s no expertise. We’re all experts at everything, right? Is this the end of expertise?

Speaker 2 [00:12:34] No, I don’t think so. But it is a different era of expertise. So let’s define what we mean by expertise. So expertise is the specific know-how or competence to do something valuable that needs to be done. So you need two things for it to really count in the labor market. One is it’s got to be useful for something else that’s valuable. Right. So it’s got to be data science not card tricks. Two, it needs to be scarce. As you just said in a second ago, if everyone’s expert, no one’s expert. Consider the job of air traffic controller and crossing guard. These are basically the same job. The goal is to prevent things from colliding with one another. But in the United States, air traffic controllers are paid more than four times what crossing guards are paid. And it’s not mysterious why, but it’s not about social value. If we had to pay crossing guards like air traffic controllers to prevent our kids from being run over on the way to school, we certainly would do that, right? So what they’re doing is valuable, but it requires no licensure or training in almost any U.S. state. Whereas becoming air traffic controller takes years of training. You have to go to an air traffic control college and then spend hundreds or thousands of hours apprenticing. Right? So that’s expertise. In some areas, I will directly eliminate the need for that expertise. Right. So I imagine actually I will do air traffic controller, most of it in ten years time. But in many other cases that’s not true. It’s going to supplement judgment. It’s not going to make it unnecessary, many of us. Right. And let’s just stipulate most people are not good at writing and never have been. Writing is hard, but most people are asked to write things. If this can enable more people to communicate more effectively, it doesn’t eliminate the need for them to communicate, but enables them to do that work more efficiently. So I think there are certain forms of expertise that will be stranded. They were valuable and now they’re not because they’re no longer scarce. Used to be to be a London taxi driver, you literally had to spend years memorizing stuff. Now it’s all done by Waze and you’re just a driver and lots of people can drive. But again, it partly depends on the vision that we pursue. This technology does not decide how it will be used, and we have lots of agency in this.

Speaker 1 [00:14:36] How do we as a society help ensure that that is the direction of travel for this new technology? So you’ve mentioned a couple of times the opportunity in health care and all the public good that would come from that. But others have noted that for all the powers of social media and all the good that it could have done in the world, what it’s really meant is a lot more amateur photographs being taken, sorted and shared.

Speaker 2 [00:15:01] That’s the good scenario.

Speaker 2 [00:15:04] Yes, and a lot of misinformation.

Speaker 1 [00:15:05] Yeah. So and that’s just the market at work I agree.

Speaker 2 [00:15:09] And I don’t think the market gets us all the way where we want to go. And I don’t think the incentives are correctly aligned. And that’s why when you ask me about Canada’s dramatic investment in this, I say I want to direct it. Right? So partly comes from government. It doesn’t mean the government has to do all the technology. It just needs to say we’re going to put out grants or contracts to develop certain things or to incentivize. It comes from universities, it comes from NGOs. You know, labour is very involved in negotiations of how technology will be used, and that affects what technologies are actually developed. It also depends on the private sector. But I also think it’s really affected by the vision of who’s doing the technology development. Your vision of what you’re trying to accomplish does affect what you do, because you may not accomplish what you set out to accomplish, but you generally won’t accomplish what you don’t set out to accomplish. I believe at the conference, you and I are both out. We saw Sebastian Thrun. He’s the originator of the Google car and self-driving vehicles, and he was sort of berating some of the people in the room saying, you know, you’re talking about get a robot to wash dishes, get a Royces. You’re thinking so small. Like, why do you want to do exactly what you do already? Think big, right? Don’t think about just putting a robot to do what a person does. Let’s just make a new capability. So I thought that was actually an inspiring point.

Speaker 1 [00:16:25] I want to ask you about the middle class, because you’ve written a lot and a lot of profound insights on what this means for the middle class. You mentioned the China shock and what that did to a lot of blue collar communities in the US, manufacturing communities. What risk is there that AI does something similarly to the middle class and particularly to white collar jobs?

Speaker 2 [00:16:49] Yeah. So the degree to which AI competes with workers is actually much more in the professions. And I actually don’t think that’s so terrible. Professionals are the highest paid people in our society, other than a few entrepreneurs, and a lot of the work they do is decision making work, and they’ve become very valuable. Understandably, computing has actually made them much more valuable because they can make better decisions. And the people who supported them, many of them in automated, but they become more and more expensive as a result. Part of the rise of inequality is the rise of the pay of college educated and post college educated workers. Now, I don’t resent anyone’s high pay. That’s great. However, their high pay is everyone else’s cost, right? It’s the cost of legal services, the cost of medical services. It’s the cost of education. And those things have had very slow productivity growth. A lot of our rising prosperity after the Second World War had to do with getting much more efficient at producing stuff. Televisions, home appliances, cars and those things are much, much cheaper than they used to be, right? You can buy televisions by the yard very cheaply. However, the cost of medical services, of legal services, of education, and we spend an enormous share of our budgets on education and health. Those have risen. They haven’t fallen because they involve a lot of expensive decision makers whose productivity isn’t rising. So if technology can one make those people more efficient, but two enable a larger set of people to do that work, it may indeed bid down the wages, a bit of very highly paid workers like myself. However, if it lowers the cost of those services and enables more people to do that viable work, that would be a very positive thing.

Speaker 1 [00:18:28] So it sounds like you’ll have a bidding up as well as a bidding down. If a lot of frontline workers, salespeople and the like become experts or have access to all the expertise that we were talking about earlier, they bring more value to their jobs.

Speaker 2 [00:18:41] That’s what I would like to see. I would like to see more valuable work done by people without college degrees. That’s the majority of working adults in our countries, and they have actually had a rough 40 years because they’ve been so pushed out of the middle and increasingly a large fraction of them do in person services, food service, cleaning, security, entertainment, recreation. Those crossing guards we were talking about earlier, that’s all valuable work, but it’s poorly paid because it’s not expert work, because it’s just abundant set of people who can do it. It doesn’t require much credentialing or training. So what would be great in a way we can know we’re succeeding is if we see more people without four year college degrees doing high value work, enabling more people to do valuable work with better tools would be a very valuable development, in a world that is seeing more than four decades of rising inequality.

Speaker 1 [00:19:35] And then all those elites that feed off that producer. Yeah, the the lawyers, the accountants not to pick on them, but those start to they don’t maybe not disappear, but start to diminish in terms of the economic rent that they gain from that front line producer.

Speaker 2 [00:19:50] Exactly. Maybe a little bit less economic rent. You know, it’s not they can become less valuable. They just might have a few more competitors.

Speaker 1 [00:19:56] This sounds almost egalitarian.

Speaker 2 [00:19:57] That would be the good scenario.

Speaker 1 [00:19:59] Technology generally doesn’t do that right. It skews it doesn’t converge.

Speaker 2 [00:20:03] Well, you know, it really varies over time, the industrial era, right from kind of 1890 through 1970, really built the world’s middle class. Certainly the middle class in the West, right, allowed people with primarily with high school educations to do a lot of very valuable work in factories and offices. Right.

Speaker 1 [00:20:21] They Left the farm to go to the city and tripled what their forbearers probably made on the farm.

Speaker 2 [00:20:27] Absolutely. And life in the factory and in the office, it’s a lot less physically demanding, a lot safer than working on a farm, a lot more economically secure. So that worked very well. In fact, that was a very good era in many ways. Not everything about it was good, but in terms of the growth of the middle class, the last four decades have not been I think computerization has really contributed to that, although some countries have handled it better than others, and Canada has had less growth of inequality and more social supports than the United States. And so now we’re another era I don’t want to forecast, say, this is what will happen. I’m really talking about what we could do, what the opportunity is. I think this technology has really different properties from conventional computing. Right. If I told you the world’s frontier technology, it’s just so amazing. But it’s not reliable with facts and it can’t do math. You would say that doesn’t sound like any frontier technology I’m familiar with. And the answer is yes, right? Because I is really quite different and used well, it can enable more and better decision making, which is really the valuable work that we do. And. Decision making. I don’t mean just sitting around making decisions, I mean in some vein where it matters.

Speaker 1 [00:21:39] When you look ahead, what might the middle class look like in an AI-enabled economy?

Speaker 2 [00:21:44] Well, if we do it well, we’ll have more people who are not from the most elite educational environments and so on. But they’ll do expert work and they’ll use better tools to do it. The world that we don’t want to live in is a world, not a world without work, but a world where everyone’s a crossing guard again, crossing guard, valuable work. But if we’re doing work that’s non-expert, and anyone who is of sound mind and body can do reasonably well, it won’t pay well. Labour’s share of national income will fall. Now, that doesn’t mean the money gets burned. It just goes to owners of capital, owners of machines. And I don’t think that’s as good a scenario. I think a world in which a lot of income comes from labour, which is the world in which we currently live, is a very good world, because in a democracy, you want people to be participants. Part of the way participation happens is people have an ownership stake in society. A lot of that comes from the ability to generate resources. They have incomes. Strong labor markets and democratic governments go well together. And without a solid middle class, it’s actually hard for democracy to work well. If all the money is in the hands of a few people who are then giving it away in the form of UBI, we’re really at their mercy.

Speaker 1 [00:22:57] What are the skills that you think people will need to hold on to and enhance to thrive in this sort of transformation?

Speaker 2 [00:23:05] So let me give two answers at a foundational level, the world is very different from when I went to high school in that period. Facts were scarce, information was scarce, and ownership of information was very key. Now, young people live in a world that’s a wash in information. Their information is unreliable. It’s very uncertain. It’s not facts, it’s conjecture, it’s conspiracy. And so a very foundational skill is to be able to work in a information rich but uncertain environment and draw valid conclusions, make the right inferences, use analytical, statistical, and scientific thinking to work in that very complex environment. But then I think at a more specific level, in almost every area of work, to be really good at, you need to develop judgment. Judgment comes from experience and from immersion, right? Most jobs that are professional jobs are jobs for which you apprentice. We just don’t call it apprenticeship. So if you’re a medical resident, if you’re an assistant professor, if you’re a junior lawyer, you’re an apprentice. And then of course, we have lots of apprenticing in the trades and so on. And so people develop that judgment when they work with AI, they’re still going to need that judgment. You want to think of AI as a quite fallible assistant. It’s always sounds confident. It’s not always right. And I don’t think that’s going to fundamentally go away. I do not think AI is going to become infallible. I do not think the hallucination problem goes away. And so it’s going to take judgment to use this tool. For example, we know my colleagues Nikhil Agarwal and Tobias Salz, among others, did a really amazing study of the use of AI by radiologists. This is reading scans as a kind of task for which I is really well-suited, because there aren’t really bright line rules. You just get the kind of the overall image and you try to focus on things and say, what do I think is likely here? So the AI is about as good as 65% of radiologists just working from scans. But it turns out that when the doctors and the radiologists work together, the doctors do worse. And the reason they do worse is because they haven’t been trained in using this tool. The tool has uncertainty, just like doctors do, and it reports uncertainty. So it will say, I’m 55% confident that this is an edema or that this is pneumonia or I’m not. So it turns out when the AI is uncertain, usually the doctors are uncertain. And when the AI is very confident, usually doctors are very confident. And what tends to happen is when the AI and the doctor are uncertain, the doctor will defer to the AI. And when the doctor and the AI are both confident, the doctor will override the AI if they differ, and neither of those appears to be the right decision. In fact, when it’s not very certain, it’s probably not very reliable. Your judgment is probably better. When it’s very confident, you should at least ask yourself why this AI that has seen millions of scans has a different opinion than you do, and at least take that into account. So that doesn’t mean that this tool can’t be helpful to radiologists, but it means it takes a training about how to use it, and b it takes judgment. So we’re going to be seeing these machines as fallible, just like any colleague you would work with. You need to know when to rely on their judgment and when to rely on your own. And so I think our interactions with the AI, we’re going to have to be very good at staying alert, remain in the loop and listening, but not uncritically.

Speaker 1 [00:26:20] David, I wonder if I can ask two quick questions in closing. One is to take you back ten years and the other to throw you forward ten years. If you go back ten years and see where we are today, what surprises you? What didn’t you get right about the world of work in the mid 2020s?

Speaker 2 [00:26:36] So two things. One is I thought the inequality would keep rising as it has been, but in fact it’s plateaued in the United States, it’s come down significantly since the pandemic. I was surprised I didn’t predict that. The others, I didn’t think that I would get as good as it has as fast as it has. I was aware of AI because I’ve been writing about technology labour markets for a long time, and I spent a lot of time with my colleagues at MIT who do AI work. And they were very divided a decade ago about where things were going. Some people were like, wow, these predictive models are really improving so fast. And others say, oh yeah, now it gets it right on average, gets it wrong in every relevant case. This is a dead end. Diminishing returns, not going anywhere. It turns out the first group was right. It was really evolving very rapidly. So that was quite surprising. And I think it’s quite destabilizing. We have less certainty about the future now than we did ten years ago, not more.

Speaker 1 [00:27:26] Take us ahead ten years. We’re in the mid 2030s. How is work? How is labour different?

Speaker 2 [00:27:33] So if we do this well, more people will find the scope of their job responsibilities expanded, not contracted. People who work in insurance. They will work with a broader variety of products. People who are doing writing, they will use better tools. People in medicine. There will be more people doing more diagnostic and support work, not simply just processing. And then if that all works out, we’ll also have higher productivity and higher productivity fixes many ills. It could enable somewhat shorter workweeks in rich countries. We have large populations who are over the age of 65, who’ve earned a retirement, and a smaller number of workers supporting them. So if we have higher productivity, it’ll be easier for us to meet that care responsibility. Now, I want to say there are a lot of things that I will do and not all of them in labour markets. So when I worry about I don’t primarily worry about elimination of jobs. Although I do worry about that. I worry much more about disinformation, weapons development, control of critical systems. But I’m a labour economist, and I think fundamentally the health of society depends on a well-functioning labour market. And I think there’s opportunity to do well in this era.

Speaker 1 [00:28:45] We’ll leave those other topics for a future episode of Disruptors. But David, this has been a great conversation. Thank you for being on Disruptors.

Speaker 2 [00:28:52] It’s a pleasure talking with you, John. Thanks for inviting me.

Speaker 1 [00:28:57] Throughout history, technology has rarely led to less jobs, although it’s certainly impacted a lot of jobs that have disappeared over time. That’s because when technology eliminates old ways of doing things, operating an elevator, for instance, it creates new opportunities. But to get it right, we need to invest in our people to enhance the value of expertise and enable valuable work. And whether you’re optimistic or pessimistic about AI, remember technology doesn’t decide how it will be used. That’s on all of us. Those listening know that AI and the promise of productivity has been a recurring theme this season. And for a country with a productivity challenge, to put it mildly, we need to figure out these challenges and turn them to Canadian opportunities. And fast. Before I let you go, one other tech note. You probably listened to our show today on your mobile device, but Disruptors is now also available as in-flight entertainment on Air Canada so you can listen in the sky as well as on the ground. I’m John Stackhouse and this is Disruptors, an RBC podcast. Talk to you soon.

Speaker 3 [00:30:11] Disruptors and RBC podcast is created by the RBC Thought Leadership Group and does not constitute a recommendation for any organization, product or service. For more disruptors content, visit our RBC.com/disruptors and leave us a five-star rating if you like our show.

This article is intended as general information only and is not to be relied upon as constituting legal, financial or other professional advice. A professional advisor should be consulted regarding your specific situation. Information presented is believed to be factual and up-to-date but we do not guarantee its accuracy and it should not be regarded as a complete analysis of the subjects discussed. All expressions of opinion reflect the judgment of the authors as of the date of publication and are subject to change. No endorsement of any third parties or their advice, opinions, information, products or services is expressly given or implied by Royal Bank of Canada or any of its affiliates.