Should we fear the robots?

Recently I was honoured to be invited to speak at the Organisation for Economic Co-operation and Development's Global Forum on Technology. This article is an expanded version of my TechTalk.

All views in this article are my own and do not represent those of the OECD.

Introduction

I’m a roboticist, but I do not believe that robots can and should be used everywhere.

That’s not to say that I don’t get excited about some of the amazing technology that’s being developed.

I do, but I also look at that amazing tech and think, how are we going to use it, should we use and how is it going to affect the humans.

In this talk I want to look at our relationship with robots, why we need them in the future, how we should use them and, if we get the relationship right, why we shouldn’t fear them.

Before I go too far, I think it's important to define what we mean by a ‘robot.’

What is a robot?

You probably should never ask a hundred roboticists what a robot is; you’ll get a hundred different answers. But, as it’s just me, I’ll try and answer it the best I can…

The term ‘robot’ is barely a hundred years old, first appearing in a play in 1920 written by the Czech writer and playwright Karel ?apek.

Derived from the work robota which translates as forced labour and more figuratively as drudgery and hard work, ?apek used it to describe machines that worked as ‘forced labour’ in a factory. Today, the word has become synonymous with machines that carry out tasks for us.

But it wasn’t until thirty years after the words appearance that the first industrial robot was commercially introduced in 1956. Yet if we look further back in time, there are examples of what we would now consider automated machines dating back to the 4th century BCE.

But what is a robot?

Well, if we ask the Internet, a robot can be either physical or a virtual software agent, more simply a ‘bot’.

But the Internet also thinks that Star Wars’ R2D2 is a real robot, so the internet might not be completely correct!

In reality, a robot is a machine that can carry tasks automatically.

I say automatically and not autonomously deliberately. An autonomous robot implies that it thinks for itself and decides what it will do and when it will do it. But not all robots are autonomous. Most repeatedly follow a set sequence of tasks that never deviates. They carry out the task automatically, but they don’t decide how to do it or when to it. But they are robots.

I one guess a confusion around robots is that they take many forms. I’ve mentioned that they can be physical of software agents. For this talk, I’m going to concentrate on physical robots.

From the humble industrial robot, programmed to simply move a box from one place to another, to complex humanoid shaped machines, robot cars, planes or even child’s toys; robots come in many shapes sizes and critically intelligence.

To some extent it could be argued that a microwave oven or dishwasher is a robot. They both carry out human tasks automatically, but we don’t consider them to be a robot. Maybe we should?

Robots are designed for the purpose they will be used for. Flying robots tend to look like helicopters, cleaning robots look like vacuum cleaners and industrial robots tend to have an ‘arm’ to mimic a human worker. Robots that interact with us tend to have faces to look like us.

For me, a machine becomes a robot when it has three core elements integrated together. It needs to have manipulation, a way to move. It needs to be able to sense it surroundings. And it needs to have a control system, a program.

Sometimes, these three things come together and end up looking like a mechanical human and perhaps that’s were our fear starts.

Especially when the robot starts to become autonomous, i.e. when it starts to think for itself and decide what to do and when to do it rather than follow a simple sequence of tasks.

But why do we fear robots so much, whether they are autonomous or not?

Fake News

Well perhaps our fears start from how robots are presented to us by the media.

If we look to science fiction, we find many instances where robots are portrayed as the aggressors, determined to destroy humans. Perhaps our greatest robot ‘enemy’ is the Terminator. In those films, robots become our overlords, destined to wipe out humanity. And there are many other sci-fi books and films that follow a similar vein.

But we can also find many other instances of robots becoming our closest allies, even friends. The roles C3P0, R2D2 and BB8 play in the Star Wars franchise perhaps epitomises this.

But these are just stories, products of creative imaginations. Are they true reflections of societies views on robots? Sci-fi isn’t necessarily always an accurate portrayal of the future or truth, but neither is our mainstream news media.

I’m sure you’ve seen the sensationalist headlines proclaiming that robots and AI will render humans obsolete within this century.

And if they don’t manage to do that, robots will take all our jobs, causing mass unemployment and displacement without any justification or historical evidence to back up these claims.

And if the robot or AI goes wrong, the media is quick to jump to the conclusion that it’s the robot’s fault, that we can’t trust the technology. In the eyes of the media, it's never the humans fault regardless of the truth.

And whilst most of these stories distort the truth, they are the stories that people see and unfortunately believe. As a newspaper editor once said, “good news doesn’t sell newspapers”.

Robots vs jobs

And as media hype grows around robots and AI, it starts to concern citizens. They start asking questions about whether their job will be automated and what policy makers are doing about it.

Some years ago, a rumour started that a senior UK Government minister was actively blocking investment in robotics and AI because one of their constituents threatened not to vote for them if they didn’t stop the “rise of the autonomous car”. The voter was a delivery driver who had read a news story about how autonomous vehicles would revolutionise the transport industry.

That story might be an urban myth, but it wouldn’t surprise me if there was some truth in it. More importantly, it highlights the fear people have that robots take jobs.

This fear of machines taking jobs first appeared during the Industrial Revolution of the 18th Century and lives on today as hard-working people fear losing their livelihoods to robots. And it would seem that no industry or job is safe.

But is that really the case?

Historical evidence doesn’t entirely support that view. Certainly, there will be disruption to people jobs. Some may unfortunately lose their jobs. But historically, evidence shows us that when an innovative technology is introduced there is rarely a mass unemployment event. On the contrary, the innovative technology more often creates new wealth, new opportunities and new jobs.

And there are countless examples across history to highlight this. The industrialisation of agriculture saw a shift of people from the land into offices and factories. The advent of the word processor saw the end of typing pools but created new opportunities in IT and service industries. Indeed, as innovative technologies replace old, new professions and industries are created.

This time around it may be different, we’ll have to wait and see. But looking back over the last 70 years of industrial robotic development, the historical picture of wealth (and jobs) being created by a new technology seems to hold true.

And for those that feel that investing in robots kills jobs, consider this. Imagine if your competitor invests in robots? History and statistical data shows that your competitor will experience substantial growth and out competes you. As a result, your business shrinks, and more jobs are lost. Why?

Simply because robots don’t destroy jobs, they enable a business, a whole economy, to grow and prosper.

Our fear shouldn’t be about robots taking jobs but should be about what will happen to jobs if we don’t embrace robots.

But we need to prepare our citizens for this. They’ll need educating in the future technologies and how they will affect them and their jobs and how to get the best out of them. There will be a need for long term support as people jobs and lives are revolutionised by robotics. People will need support as they adjust to this new paradigm.

And equally importantly, the narrative around this increase in robot workers needs to be managed sensitively. The role the media plays in fuelling the fear people have around robotics and automation needs to be tackled head on and changed.

Unfortunately, the robot industry isn’t great at telling a good story. Too many times, the story is about the technology and not what the robot is doing, or more importantly how it is making the life of an individual safer, more fulfilled or simply better. The industry needs to change its narrative from how clever the science and engineering is, to how the robots will make people's lives better.

“Trust me, I’m a robot”

We can educate people about robots and their benefits, but how do we get them to trust them How can we trust a system that we are not in control of?

Human beings are at are core, ‘control freaks’. We want everything to be under our control; to be the way we think it should be.

We can see this in the way we have manipulated mother nature to our needs. By cultivating land for crops, domesticating wild animals and building cities, we have controlled our environment. And if we can’t control it, we react with a fear or flight response. We stay and fight to control it or we run away from it.

I think it’s entirely natural that humans will want to control the machines that surround us and feel fearful when they begin to act autonomously. My belief is that humans will always have the final control over the machines. There are many that don’t believe this. They believe that robots will take over every aspect of our lives and eventually we will lose control of them. I hope they’re wrong.

I still want to control the robots that I work with, but I do want them to think for themselves and act autonomously. But for many, that is a scary proposition. They are fearful of a robot that is thinking for itself. How can they trust that it has our best interests at mind?

Not trusting the robot is one thing, but how much trust do any of us have in those building and controlling these systems? How transparent are they in explaining how the system works and its intentions?

I’m not saying that everyone developing an autonomous system has bad intentions, far from it. ?But how can we be sure? How can policy makers and citizens trust the intentions of developers outside of their regulatory reach? What damage could foreign actors to with autonomous robotics?

This fear is often worsened because current AI (and the robotics it controls) are not infallible and every time an autonomous system fails, its headline news and our trust in the robot and AI is eroded.

Tackling this trust issue involves educating citizens, changing the narrative around robots but also greater regulation over how they’re developed, how they operate and who operates them. It's far from an easy problem to solve, but it will be crucial to our use of robots in the future.

When and Where

One thing we can do to alleviate this mistrust in robots is to use them where it’s appropriate.

At the start of this article, I talked about the origins of the world robot. That it comes from a Slavic word and figuratively translates as forced labour.

Unfortunately, we’ve often used forced human labour to carry out the dull, dirty and dangerous tasks that the others did not want to do. But that’s exactly why robots were developed, to carry out those tasks that we didn’t want to do.

The first industrial robots were used in automotive factories carrying out tasks that humans found dull or repetitive, where often in dirty, hazardous environments and in many cases were dangerous to their health. Even today, many robots carry out tasks that humans have deemed to be dull, dirty and dangerous. That could be packing boxes, lifting heavy objects or working in hazardous environments like nuclear power plants or similar.

Yet we’re also seeing robots start to permeate into our everyday lives. Is vacuuming dangerous, probably not but it may be dirty, and some may find it dull. So, a robot could do that task? How many other every day ‘human’ tasks could a robot do?

Some think that robot could or should take over many of our tasks, allowing us more free time to do whatever we please. I’m not sure it’s that simple.

Yes, we absolutely should use robots for tasks that are dull, dirty and dangerous and in environments that are not suitable for humans or are too demeaning. But should they do everything.

Do we need a robot to wash our dishes? We may find it a dull job to do, but do we need a sophisticated humanoid type of robot to do it? That robot might be quicker or even better than a human at the task, but so is a dishwasher commonly found in most kitchens! This might be a case where using a robot is not necessary.

And whilst robots (and AI) are becoming more intelligent, more conscious even, will we ever be able to replicate that intangible thing that’s makes us human. A robot is after all, a machine controlled by algorithms. It might be able to learn using sophisticated AI but will it ever be able to capture the true idiosyncrasies of being human?

So where do we use robots?

So perhaps we now need to think very carefully about our relationship with robots. What do we want them to do for us?

Are they our ‘slaves’ of the future, destined to be downtrodden doing the things we don’t want to.

Do we want them to look and act like us? It’s a very real possibility and one that can excite and scare in equal measures. But here, we run the risk of the “uncanny valley” affect? That’s where the more human we make a robot look, the more we find them creepy and distrustful.

Perhaps if we want to overcome our fear of robots, we need to recognise that they are machines and stop making them look like humans?

Perhaps, but our human bodies have adapted to suit the environments we live in and so maybe our robots should evolve to look more like us. It’s a compelling argument. Make the robot look like us as we have shaped the world to suit the human form.

But the human body is a frail thing, a ‘jack of all trades’ to suit our world. Our bodies don’t survive well in harsh, dangerous, dirty environments. Our minds prefer to be stimulated to allow us to thrive, create and well, be human.

So instead of the headlong rush into making robots that are more like us, perhaps we should start making robots that are specifically adapted to do the things that humans aren’t good at in the environments that we don’t do well in. That may well mean that they don’t have arms or legs or a face but are designed to get to specific places and to do specific jobs.

For me, right now the robot industry is putting too much focus on making robots human but forgetting that robots are there to make our lives better, not replace us. Robots don't have to look like humans to achieve that.

?

Conclusion

We need robots that we can live alongside us that we are not afraid of because we need robots in our future.

Robots can survive in the harshest of environments. They can clean up toxic spills, survive harsh nuclear environments and be unaffected by pathogens. We will need robots to clean up our environment, support our ageing populations, carry out those dirty, dull, dangerous tasks and in doing so, improve the quality of life for so many people.

Some might look and act like us, but I think most will not and should not.

When we’re thinking about how robots fit into our future, I think we need to consider some key points:

·???????? We must choose how we apply robots carefully.

·???????? We must recognise that they are not human, they are machines.

·???????? We must use them when and where it’s appropriate to.

·???????? We must understand their development, educate people on the benefits and uses and most importantly…

·???????? Understand when a robot cannot replace that unique thing that a human is.

We should not be afraid of a future world with robots beside us. But we must prepare ourselves for that future and understand how and when we should use a robot.

?

?