This is adapted from a speech I gave to robotics and automation engineers at Schunk Expert Days in Odense, Denmark, in early 2019.
I’m going to start by asking you a question.
You could be anything you wanted to be. You could have been a farmer, a doctor, an artist … you could have worked in retail, or many other things. The question I have for you is: Why did you choose to do what you do today?
Why did you choose to become who you are?
I don’t know you, but I assume many of you are like my oldest son. Ethan is currently a third year engineering student at the University of British Columbia, in Vancouver, Canada. He’s studying mechatronix and doing lots of projects … a solar powered boat, a mechanical clock with a stepper motor he ripped out of a scanner my dad used to own, a drone project for a competition in Shanghai, and other things. Since he was just a few years old, he’s always been the kind of kid that wanted to know how things work and how they’re built. What are the components? What are the pieces? How are they put together? How can I fix it? Or more likely, especially when he was young: How can I break it?
I kind of have a feeling that many of you are not too dissimilar from that. And that’s one of the reasons why you chose this particular area that you’re working in right.
I have a second question for you.
Do you have any clue at all — whatsoever — how critically important you are for the future of humanity and the future of this planet?
Do you have any clue how important you are?
One thing that’s really obvious is that things are changing at a faster and faster rate. You don’t have to believe in the concept of the singularity to see it and to feel it.
Imagine a simple chart. Then plot the rate of change on the vertical axis and time on the horizontal axis. For much of human history, it’s been a very uninteresting graph, just continuing on straight and level for thousands of years. Some hundreds of years ago that line started go up. And now it’s going up faster and faster.
The concept of the singularity is that at some point, the rate of change becomes so immense, that curve almost becomes a straight line up. This is fueled by artificial intelligence, robotics, and automation. Well, you don’t have to believe in that concept — that we will actually come to a singularity — to see that the rate of change we have right now, just over about the past decade, is immense.
Mobile blew up from almost nothing to the single most ground-breaking communications channel or media that we’ve ever seen in the course of about a decade. Over a similar time frame, we’ve seen IoT — the Internet of Things — go from not much more than a concept to about seven and a half billion connected smart devices last year, and still exploding to 25 billion by 2025.
In 2012, we had one social platform that connected a billion people. Today, we have six. In a couple years, that will likely be eight or nine.
All of this has contributed to an avalanche of additional change.
One is “smart matter.”
Smart matter is what I call the addition of sensors, motors, chips, and radios to every thing, every piece of matter, every atom around us. Think of an Amazon Echo, for instance, it’s got sensors: it’s listening. It’s got motors: it has a speaker functionality, which moves air to speak, to play music. It’s got radios to communicate via WiFi. And it’s got chips to manage and control all its capabilities.
Another example: you can buy a smart toilet today that has Alexa built in.
So you can walk up and the toilet will open, you can leave and the toilet will close and flush. And, if you are so inclined, you can have music playing while you are sitting on the device, thanks to Amazon Music.
We also have smart glasses as well. For instance, HoloLens 2 was just released a few days ago. And of course, nobody in his or her right mind is going to walk down the street wearing it even though it’s slightly smaller, slightly better than the HoloLens 1, but in five to seven years, we’ll start to see consumer adoption of smart glasses that do basically what this does, but also paint default reality with augmented information and do many other things as well.
Of course, there’s other things that are changed as a result of all those things that I talked about.
We’ve also seen the rise of surveillance capitalism, right? The Facebook business model … if you don’t know what the product is, you are the product. You’re trading data for access.
Free global communications?
Think about it: you come to somebody in 1975, and you say, you know, there’s this concept of a global communications platform, and you can be connected to anybody anywhere on the world in full audio and video at any time for free. They’d have no concept of that because the only concept of communication they have is something that they pay for, and they pay more for long distance.
Anybody remember long distance here? Yeah, exactly.
Some are old enough to remember. But many of us are young enough that we will never use a communication medium that differentiates between short and long distance.
But that’s one of the consequences of some of the change that we’re seeing. We’re also seeing cultural changes: renewed clashes between globalism and nationalism. As we’ve had these billion-person social platforms, we’ve come to grow more and more into our own reality bubbles.
We see more of what we engage with, and therefore believe more intensely the things that we already believed in the first place. Somebody once said that smart people have strong opinions loosely held.
Strong opinions, you really believe what you believe. Loosely: I can change my opinion, based on evidence.
We’re coming more and more to the point where for many people have strong opinions, tightly gripped, and almost impossible to change … even as we have macro level changes that are occurring that have significant urgency, such as environment versus growth.
Question: Is all growth bad?
Can we have growth without environmental damage? What does growth even mean in an enclosed system? How do we grow economically in a financial paradigm that cannot continue to grow forever? And, can we find ways to grow while also making the world a healthier place?
So these are major changes that we’re experiencing and that we face.
But I’m going to argue that they’re not the most significant change that you and I are going to face over the course of our lifetimes. There’s a much more significant change coming: the end of work.
I’m talking about the application of automation, robotics, and artificial intelligence to the things that we currently do to create value … to make things that we need, and to make a living.
According to a study that came out of Oxford University about a year ago, 20-70% of the jobs that currently exist on the planet are at risk of being automated via some combination of technologies. Only 5% of the jobs are actually “safe.”
Well, let’s examine that. And let’s be a little critical about that. We know, for instance, some jobs will never be automated. For example, Elon Musk was very open about the ramp to production of the Tesla Model 3. He said, we are going to build the machine that makes the machine. And it was very, very obvious that Tesla wanted to build extreme amounts of automation in the process.
But that turned into production hell: the Tesla Model 3 had very low production numbers for a very long time. And what Elon Musk had to do eventually is re-admit people into the equation. And he was actually surprised by the result. People are actually really good at this stuff, he said.
That was interesting.
Should everything be automated?
If we look at what we eat, perhaps food that a high end gourmet chef creates … do we really want that automated? Do we really want the art of cooking to be something that is done by a smart system? What about counseling and psychiatric help? Is that something that you want to do: talk to a robot about your problems and your challenges? And one of the things we love about athletics is the amazing abilities that some of us possess that transcend the limits that most of us have. Do we really want to watch robots play sports? What about parenting — is that something we can replace with an automated system or robot?
Well, let’s look closer.
Actually, yes, Tesla did have production hell. They did have to reintroduce people back into the equation. But is there anybody here who would be willing to say if we look at Tesla 20 years from now it’s not going to be more automated then it is right now?
Of course it will be.
Sure, it’s challenging to get there. It’s not easy. It’s hard to do. But it’s going to happen.
Everything that can be automated, will be automated.
Look at McDonald’s. Most of what we eat every day is not amazing. We had a great dinner here, by the way, and thank you so much for that. But most of the food we eat every day is not super high end. I haven’t been here in Denmark before, but you have a lot of McDonalds restaurants here. I think I’ve seen about 500 already in the short period of time I’ve been here.
Automation is going to hit the food industry hard. In North America, you often see order screens in McDonald’s, which of course is potentially taking the job of somebody who would take an order from you. I’m sure everybody is aware there are burger-making robots. You can go to a restaurant in San Francisco and eat a burger that a robot has made for you. Do we really think that in 10 years or 15 years … maybe even five years … McDonald’s is not going to adopt that kind of technology?
Our Big Macs will certainly be made by robots.
What about mental health?
Some of the first iterations of what we currently call chatbots were made in the 70s and 80s. They were essentially psychological systems which somewhat cleverly parroted back to you what you had said … which of course is a psychiatric technique. Today you can get an app called The Robot on iPhone right now and it’ll help you with your challenges and problems. I do not recommend it, nor have I personally tried it.
You see the robotic World Cup of what we call soccer in North America but you call football here. I don’t think most people would pay money to watch them — except to see the spectacle. But where’s that going to be in five years?
In fact, there’s been serious discussion in the United States about the NFL thanks to the concussion crisis. They’ve had to pay billions of dollars in compensation to players. And there’s been serious discussion that at some point, that game will be played by robots. I don’t know if that’ll happen. I tend to like the human component of these things.
But that’s interesting, right?
And there’s concussion issues in ice hockey, my favorite game, of course. (I’m Canadian, how could it not be … it’s basically required). But there’s also issues in soccer/football since players have to head the ball repeatedly.
And of course, we’re not even talking about eSports. People are making millions of dollars as eSports champions right now, and many are watching and following them on Twitch and other live streaming platforms. That’s not exactly robotic. But it’s interesting and new.
Automated health care is a big opportunity.
Perhaps we’re not going to automate our mental health care to robots. But there’s huge opportunity in physical health care. In Japan, for instance, elder care is a big, big deal. There’s a lot of different needs that the elderly have in Japan as a rapidly aging society, and they’re looking to robots to help with it, from companionship to movement, lifting, carrying.
This is probably a picture of where many of the rich modern industrial nations are going to be soon, as the birth rate continues to decline. And there are not enough young people to take care of the old people.
So you’ll need robots that help our elders. Some of them are, are just mobility robots helping you get around move, some of them are companionship robots. So there is actually a lot of growth that’s happening there.
And, of course, industry is a major growth area for robotics.
This is one of those areas where you are experts: in the recent five year period from 2012 to 2017, the worldwide supply of industrial robots doubled. That’s something you know a lot more about than I. But as somebody who watches change, I look for doubling affects over short periods of time, because they end up being super powerful.
Exponential change …
And somehow our human brain doesn’t actually compute those very well. Everybody has heard the story of the lily pads that doubled every day for 30 days, right?
So you walk by the pond every single day, and most days you don’t really notice anything at all … there’s a couple of lily pads. So what? And then another day, the lily pads cover half of the pond. When did that happen? The day before day 29, right?
To human perception, it was nothing and all sudden, this doubling effect changes everything. That kind of change is disruptive, that change doesn’t compute easily in our brains.
It’s not just about industrial robots. It’s a lot of things.
Look at law. There are expert systems, AI-delivered systems for contract compliance assessment. There was a test done about six months ago on contract compliance that pitted human lawyers against the AI. The result: the very best of the human lawyers perform better than the AI. The AI performed at about the 87% range. That’s above most of human lawyers … but the AI did it in about three minutes.
And the best of the human lawyers took over an hour.
That’s just one area. Accounting, obviously, is following rules for what you need to do with your money and account for your assets and things like that. Lots of smart systems are also being built for more creative verticals, like marketing. Marketing has traditionally been a very, very artsy thing. But there are nowsystems that you can give instructions right now about what you’re selling, what assets you have, and what you want to achieve, and it will go out and spend according to your goals. The system will determine where it’s going to allocate money, where it’s going to market, where it’s going to advertise, what it’s going to do.
You can watch over it and you can set some parameters for what not to do.
What about health care?
Health care is a global disaster. Ideally, we need much more automation even in rich industrialized nations. There are still many people who are underserved medically … especially true in the United States are many people can’t afford it, since it’s a for-profit system. But the needs are incredible in a place like India where nearly a billion people don’t have adequate access to medical care. Half of Chinese people cannot access healthcare. Many nations in Africa, for instance, do not have sufficient healthcare.
So we need automation and we need intelligence in health care.
In fact, there’s a smart system that iFlyTek, one of the top 15 Chinese companies, has developed. And it has passed the Chinese medical exam. There’s a ton of room to grow here, of course. But we need more AI and more automation if we want to have a hope of delivering essential services to everyone.
There will, of course, be an impact on jobs. In fact, perhaps 20 to 70% of our jobs are at risk over the next few decades.
Of course, many economists, futurists, technologists look at that and point to the past. Every time we’ve had technological disruption in the past, we’ve created new types of jobs that we didn’t even know existed … that didn’t exist or couldn’t have existed at a previous level of technological innovation.
Absolutely, we always have.
But just maybe, the past is not a perfect predictor of the future. If it was, I think we could avoid a lot of problems. However, I think that today’s changes are qualitatively different than many of the technological changes that have happened in the past.
Autonomous machines, thinking robots, and smart systems that operate on their own within certain defined spheres (which are rapidly expanding) is a qualitative difference from what we’ve seen in the past. If that’s true, that’s a problem because we don’t need 70% of jobs disappear for us to have a big problem as a society. I wonder if anybody here knows the percentage of jobs were lost in the Great Depression.
In the United States, for instance, one in four jobs were lost in the 1930s during the Great Depression. And that caused massive cultural, political, and economic change with lasting consequences.
So losing 70% of our jobs would be catastrophic.
Also, people have a need to be needed, people have a need to contribute. People need a living wage, and people need a place in society. So this is a real challenge. This is a real problem.
On the other hand, let’s be honest, many jobs — maybe even most jobs — really suck.
They’re not jobs that demand your full creativity, they are not jobs that use your full human potential. They’re not jobs that require you to be fully engaged in what you’re doing. These types of jobs might be 30% of the jobs in a rich industrialized nation. It might be something like 70% in a poor nation.
But maybe there’s a solution. And maybe it’s not just for jobs. And maybe it’s not just for some of the major problems we talked about, especially the environmental problems that we face.
Maybe it’s for both because right now we face serious and massive global changes. I’m talking about planet-scale problems like plastic in the oceans, like transitioning to 100% renewable energy, like carbon sequestration (removing carbon out of the atmosphere), like building clean, liveable safe homes for 10 billion people with less environmental impact. And we currently have reclamation projects on a global scale: healing our lakes and rivers. In the US alone, the Environmental Protection Agency has over 30 sites right now that are not being fixed the way that they should be: toxic sludge, nuclear waste, and nasty things like that.
Ultimately, these are engineering problems.
And I believe they have engineering solutions.
I wonder if you agree with me.
Remember, I asked you earlier at the start of this presentation why you became what you are. Why you do what you do. And I asked if you knew how critically important you are.
Globally, there’s a very small number of people who can meet this challenge, who can fix this mess. And the only possible solutions involve much more technology that we have right now. A ton of innovation. Much more automation. Much more work on human-robot cooperation. Much more robot-robot cooperation.
Solving some of these massive challenges will require multiple different types of robots working together. Each coming in at the right time, the right place to make sure the entire job gets done, much like humans with different skill sets work today. And we’ll need human to robot cooperation: working together not just in immediate space and real time, but also working together in cooperation over time and distance … and getting multiple robots in major projects working in semi-autonomous to autonomous ways.
We need more robots, more AI
Ultimately, what we need as a world is more robots and more automation. We desperately need more robots and more automation to tackle our huge problems and huge challenges … ones that we can’t actually fix at our current level of capability.
We saw today in many of the talks that robots are mostly focused on jobs that humans can already do. Tomorrow, we need many more robots that can do jobs that humans can’t do.
Think about Fukushima, which has been leaking radioactive seawater for years.
Just this past week, Japan finally managed to get a robot into one of the most radioactive areas and pick up one of the pieces of the original core of the reactor. But Japan has “killed” dozens of robots with increasingly higher levels of shielding that have been trying to go in there safely, and extract some of the material so they can continue the cleanup.
Other work that we need robots to do that humans can’t do is ocean cleanup. For instance, the collection of all the plastic in our ocean. That’s exactly the kind of thing that I’m talking about: planet-scale challenges, and solving them via automation and robotics.
In the US there’s very little recycling and all the garbage goes to the landfill undifferentiated. Only a few states have some level of recycling like we have in place in Canada. We have quite a bit of sorting, different pickup for recyclable material, etc. I assume you have that in Germany, and many other places as well. Denmark, and other EU nations, Switzerland and others like that. But still, it’s a real challenge. And there’s people involved in that process right now. We can’t scale it the way we need to scale it.
We need automation and robotics to sort and recycle.
We even need it in space.
It’s getting harder to send a satellite out to space now. Our modern economy depends on satellite communications and technology. We’ve set up so many satellites that they’re fillin gup orbital spots. And then there’s space junk: we using explosive bolts for releasing payloads, sometimes destroying satellites. We’ve abandoned satellites, which have the risk of colliding with others and forming massive debris fields.
And perhaps you remember a few years ago that China destroyed a satellite with a missile? It was a massive international incident, and now there are thousands of tiny pieces whizzing around at 20,000 kilometers an hour or more. Sounds like we need space garbage collection, and it’s a great job for a smart robot.
One of the things we’ll have to do is to reinvent the definition of work.
If our definition of work was us doing things manually, that’s going to change when we have more robots working with us, right?
We also need a new definition of tools.
What does a tool do? Is it something as simple as a thing I can use with my body in person, or is it something as complex as something I can use at a distance, maybe in VR, and maybe with multiple other automated systems.
We also need to redefine and re-understand the economics of value. All this work will need to be funded. Today, robots do what humans could do because there’s value created in those processes, and they are creating products or providing services that other humans will pay for. So there’s economic value there.
But now, we need to build economic value in areas that are currently underserved, that need to be solved and fixed. And many of these areas that don’t currently have an economic value. There’s no price tag on cleaning up the plastic in the ocean. There’s no market where someone earns money because she or he cleaned up a gigaton of ocean debris. But we need that kind of market.
These are engineering problems
These are hard problems. But they’re also engineering problems. And that’s what’s interesting.
The other thing that’s notable here is that time is short. We’re gaining new robotics capabilities and technologies quickly … in fact, not a moment too soon. Really we could use much right now, because this climate change crisis is urgent. In fact, some scientists are getting more and more concerned that we’ve already past the tipping point. And we’re at a level at which we’re going to see significant ocean level increases within the next few decades.
So there’s a lot of fear of the future right now. That’s not always been the case, certainly in America, the “land of progress” in the 40s and 50s and 60s. Americans used to have a feeling that progress was inevitable, that we were moving upwards as a people and as a species. And I think some of that was felt in Europe as well, probably with a little darker tinge. But there’s now a lot of fear globally about the future: what does it look like?
Change is scaring us now.
I think engineers are by definition hopeful as well as skeptical. But as someone hopeful, my challenge for you is that as you build the technology, don’t forget the sociology. As you build the tools, remember the humans who are going to use them. And as you build the engines, remember that world that has to sustain them.
Your job is to create the future.
You’re building things that don’t exist. You’re making things that haven’t been built before. That’s your job.
But a job isn’t necessarily a mission, right? “Mission” is bigger than “job.” A mission goes deeper. A mission matters more … you work harder for a missio then you work for a job.
I’m thinking maybe that mission could be using what you’re building and using what you’re able to create to invent a future in which everyone on the planet can have clean air. Can have fresh water. Can actually have a meaningful role in society and be paid for the value they deliver. In which everyone, even if they’re working with automated systems, with artificial intelligence, with robots … has the ability to feel some hope even in a world of constant change and challenge.
The amazing part of that is that if you choose to make this your mission, you get to accomplish that through what you love most: building things, solving problems, making robots, making automated systems, designing self-learning systems, and all the other things you do.
And that is why I asked you a question when I started this presentation.
Do you have any clue how critical the tiny segment of humanity that is engineers is? Do you have any clue how important you actually are? I trust you do now.
Thank you very much.