Are you going lose your job to a work robot?
By now most of us know that we’re on the cusp of a Fourth Industrial Revolution. This newest revolution is brought to us thanks to advancements in artificial intelligence, robotics, nanotechnology, 3D printing, genetics and biotechnology. These advancements are accelerating because they are all building on, and amplifying one another. Progress in these areas is now exponential, and the work robot is becoming a reality.
People are worried that they will be replaced in their jobs by a ‘work robot’ and they should be. Even Mark Carney, the Governor of the Bank of England warned that up to 15 million jobs are under threat from automation.
By the way, that’s c.48% of the current workforce population!
Most people think that the kind of jobs at risk will be blue collar, manual, semi-skilled repetitive jobs. This is not the case, intelligent automation will affect all jobs. It will do this by either completely replacing humans or reducing the number of humans required to perform tasks. It’s what the military call the force multiplier effect. On bullet can kill one person (unless you’re really unlucky), but a grenade can kill many, maybe 10, so a grenade multiplies your force by a factor of 10. Intelligent Automation will be a force multiplier, reducing the human inputs required (in some cases to zero) to get the same end result. People are defiantly worried and Google hints at the current wave of concern. Type ‘will robots’ into the search bar and you can see the most common searches.
People are defiantly worried and Google hints at the current wave of concern. Type ‘will robots’ into the search bar and you can see the most common searches.
This new revolution will be more impactful and more comprehensive than revolutions 1 to 3. We already know it will fundamentally affect the way we work, or in some cases, whether we’ll work at all or work alongside a work robot,
The challenges of climate change are driving advancements in farming technology, supply chain management, smart homes and cities. Everyone can partake in capitalism now as the emergence of the sharing economy allows people to monetise their possessions such as empty rooms (Airbnb) and cars (Uber). There is absolutely no doubt that, as with previous technological advancements, this new industrial revolution will have a profound effect on jobs, and the nature of work.
The Work Robot: What does this mean for you? Are you vulnerable? Hopefully, I can help you answer these questions.
Firstly, what do I mean by the phrase work robot? I’ll admit I’m being a little gratuitous with the use of the word in the title. What I really mean is ‘Intelligent Automation’. Emerging technologies and solution’s use AI to a greater or lesser extent and so threaten both white and blue collar jobs across all industries. As I’ve already mentioned there are many technological advances that are combining to make our lives easier or address specific problems that we face.
Fig 1. lists the technologies that make up Artificial Intelligence, all of which are advancing at pace. To cut a long story very short, advancements computing power and the way that technology can interact with humans and the wider environment is making widespread intelligent automation a certainty. That being the case, the question arises “what is ‘intelligent automation’ good at”? Most people think that their job is beyond that of a robot, but the bad news is that they are probably wrong.
Professor Vasant Dhar has modelled a potential evaluation matrix which gives us a clue as to which jobs might be vulnerable. His model is a variation on the standard risk management matrix of impact ‘v’ probability of a risk. In other words, we should focus more on risks that, should they transpire, are the most damaging and most likely. His model consists of 2 axes:
- Predictability – is the task codifiable and are all the cases that a human might be presented with predictable?
- Risk of failure – what are the consequence of human/intelligent automation getting it wrong?
Some environments are highly predictable. In deciding whether a job role is within the scope for potential intelligent automation we need to understand the level of predictability. Considerations include the degree of definable process and rules that the job entails. Is the environment within which the job role takes place stable and defined? Are the inputs and outputs to the job measurable and defined? Can the current technology manage the situation to a predictable outcome? Is the consequence of a human decision fully definable? We know this much, the work robot is very predictable, but the environment that we introduce them into must also be relatively predictable.
In terms of risk, we can look at the likely impact of a mistake made by either a human or a machine. A worst-case scenario might be a mistake which leads to a catastrophic event. Such events would lead to the loss of life, whether it be one life or many. Other risks include financial loss and corporate risks such as risk to reputation.
Fig 2 summarises this view. The jobs added seek to describe the balance between risk and predictability. In financial trading, where we are currently supported by predictive technologies, the difference between long and short term financial trading is that the longer the trade the more environmental unknowns there are and the higher the financial risk. At the other end of the spectrum, driverless cars operate in a highly predictable environment, roads are tangible and solid, have rules and generally other cars behave predictably. However, the risk of failure for both the human and machine driver is potentially catastrophic, with lives at risk. For cataract surgery, more precision is achieved by a robot and the risk is moderate, given that the Patients eyesight is threatened anyway.
Current advancements are producing commercialised solutions across a wide range of tasks that are currently completed by humans. In robotics and automation, we can all probably visualise a factory floor where robotic arms and machines construct other machines or package items. There have been recent advancements in this area and we can evidence that at Foxconn, the Taiwanese firm that produces iPhones. It has recently eradicated 60 thousand jobs having embraced robotic automation where repetitive precision is required. That advancement sounds intuitive because we are all aware that robotics has been in the manufacturing space for 20 years. The advancement here is that extreme precision is now achievable.
However, robotics is now entering areas that are less intuitive. In the fast food industry, robotic burger joints are popping up. These are operations where kiosks receive your order (& will know your previous preferences) and machines make the burger precisely as you like it.
Humans are judgemental
There is evidence that automating customer service not only drives down service costs but raises the cost of the average bill. This occurs because individuals don’t like being judged and so will order more. Automation avoids the negative judgements that occur when dealing with humans.
In the UK, the biggest fast food outlet is McDonald’s with c.1300 restaurants. If we assume that each restaurant employs at least 15 full-time equivalents and a 50% impact on staff we are looking at the loss of 10,000 jobs, for just one chain. Imagine a similar read across to the 350 thousand employees in the UK fast food industry, the impact is huge.
So, if you think about your own job role and map it against the two criteria perhaps you can visualise where you are in the matrix. That’s fine, and you should do that, the trouble is that the points on the matrix are changing. The most obvious reason for change is against the horizontal X – axis because AI and predictive analytics is improving to provide much more predictability. So, if your job role falls to the left of the human/intelligent automation line, it might not stay there, because the line is moving towards you, and fast.
Are you at risk?
Your job role might also move against the vertical y-axis. Further regulation, bringing about increased liability will move a job role vertically upwards, i.e. increased risk, whereas a reduced regulatory burden will move it in the other direction. In this way, legal jurisdictions may seek to protect jobs. Drivers, for example, will be extinct in the coming 20 years. If governments seek a regulatory choking off of the introduction of driverless cars and trucks, or triple insurance premiums, then drivers may be safe for now.
So, here are a few job roles that I believe are vulnerable:
- Doctors ‘v’ Nurses – in certain areas of diagnosis, such as diabetes, current technology throws up an unacceptable number of false positives. This error is reducing however. In the area of diagnosis, doctors will be replaced because the task is wholly predictable and codifiable. At risk in this space are GPs, of which there are 43 thousand in the UK alone. However, nurses are great for administering remedies and dealing with people. It’s easy therefore to visualise a reduction in the number GPs surgeries, diagnosis and minor treatments occurring at home or in the workplace and immediate referrals to specialists should this be required;
- Pilots ‘v’ Cabin Crew – much of a pilot’s task is currently automated/assisted. The balance will increase in favour of intelligent automation. Pilots are a relatively expensive human resource (per hour of flight) and they are prone to human error, whereas the pilotless plane will be cheaper per hour, and will be less prone to mistakes. Cabin crew one the other hand, may well be safe for the moment.
- Teachers – are also vulnerable. There’s no doubt that teachers are great at dealing with young people. The consequence of a teacher/machine passing on the odd bit of wrong information is, in the round, insignificant. However, there are different challenges for different education sectors. Imagine a school where all the teachers have the characteristics of the top ten human teachers in the country. Jill Watson had been supporting students at Georgia Institute of Technology for 5 months before they were told they were actually dealing with IBMs Watson system.
- Accountants & Lawyers – operate in a highly codified environment and there are already elements of these professions that are being automated. Algorithms now overturn minor fines such as parking, where previously an expensive solicitor would be required. AI is now also supporting the discovery of precedence and the finer points of law, tasks which were previously completed by junior lawyers.
So, if you want to know if you’re at risk of being replaced at work by a robot firstly ask yourself what would happen if you made a mistake because that’s the same consequence for intelligent automation. Does someone die? Is there a financial risk? If so how much? Also, ask yourself how much value do you bring to the environment of the tasks you perform. If the environment is completely unpredictable chances are you bring lots of value and it’ll be a while before you are replaced. Notwithstanding the predictability element do those that currently interact with you value dealing face to face? In some areas of customer service, it’s been proven that people like the fact that a human is servicing them.