By James P. Gunderson PhD and Louise F. Gunderson PhD.
Is your job as a marketing expert on the chopping block?
“Kiss your job goodbye. You have been replaced by a computer.” At least, that is what the news would have you believe. The headlines are full of dire news stories about one occupation after another becoming automated. From the mid 1900’s until the beginning of the 21st century it was primarily manual workers that were at risk, the blue collar gals and guys. First it was assembly workers replaced by machines, printing presses went digital and thousands of typesetters were out of work, and then typists were replaced by word processing systems, and so it went. However, those of us who did ‘brain work’: the information workers were safe. Until recently, anyway. But now advances in artificial intelligence are putting whole new classes of occupations are risk. Yours could be next.
But, don’t start cleaning out you desk quite yet, the robo-devil is in the details. To get a better understanding on what has changed and what to expect, we need to dive into a little history, take a side trip into economics, and then peel back the curtain on A.I. While we are doing this we will learn a little about IBM’s Watson, peek under the hood of self-driving cars, and delve into the creative processes of the human brain. Then we can really ask the question “Will I be replaced by a robot?”
The history of automation
Automation is the process of replacing a person with a machine. It has been going on for hundreds of years, and has been a dream for thousands more. We can go back to the early Greeks and read stories of cocktail parties on Mount Olympus where the drinks and canapes were not passed around by faceless servers in formal wear. Nope, these crafty story-tellers told of magical machines that moved from place to place at the party carrying food and drink for the party-goers. A great image, and one that has taken several thousand years to make real. But along the way so many other jobs went the way of the dodo.
The work by Darren Acemoglu and David Autor of MIT suggests that the process of automating jobs (either robots for physical tasks or artificial intelligence software for cognitive tasks) is following a trajectory. We have already seen the Manual Routine tasks fall to automation. We are watching the Cognitive Routine tasks being nibbled away for humans. But what is next?
Surprisingly, it may not be what you think. It is a challenge of Brains versus Brawn, and Brawn seems to be lagging behind.
Brains versus Brawn in the job market
There seems to be a natural progression in the stages that a country goes through as it grows. Often there is an initial stage of exporting raw materials. Stage one is selling off trees or ores or other natural resources to develop an more sophisticated infrastructure. Then there is a stage of manufacturing, producing finished goods for both internal and external consumption. This is often followed by a service-based economy using the knowledge and skills of the workers to provide value. At each of these stages the occupational skills required to be successful change as well.
Human labor is critical in these early stages for cutting down trees, digging out ore, and so on. In the manufacturing stage the critical skills include process knowledge but also require hours of dull repetitive work. In the later stages the focus shifts to cognitive skills. Brain work becomes prized over manual labor, knowledge trumps brute force.
At each of these stages, the cost of having people do the work is often a controlling factor on a business’s success. Businesses that can control costs grow; businesses that can under cut competitor’s prices grow rapidly. As an example, by installing advanced assembly line technology in the early 1900’s Henry Ford could both lower costs and pay higher wages for workers that could adapt to the new working conditions. He used automation of the physical process (the brawn) of assembling cars to become the dominant player in the automotive industry. His breakthrough was replacing human labor with large industrial assembly line devices.
This approach to automation became the most successful model for car manufacturers and many other industries through the first half of the 1900’s. It also put a large number of human workers out of work, if they couldn’t adapt. The U.S. saw a second wave of automation in the second half of the 20th century as these new-fangled devices called computers entered the work force. In fact, until the 1960’s the term “computer” was an occupation not a device. Successful companies hired hundreds of ‘computers’ to do information processing to keep their businesses thriving. But a human ‘computer’ was expensive, error prone, and demanded things like paid time off and medical benefits. So the economics of business led to more people being replaced by machines. Best of all, from the business’s perspective, these information processing machines were much cheaper than those huge industrial automation devices. So the benefit of replacing a human with a machine was much higher for information workers. This is even more true today, replacing information workers is the next threat on the horizon.
Think about it – buying and running a self-driving car costs tens of thousands of dollars per vehicle, in an economy where a human driver gets paid close to minimum wage. The computers needed to replace an accountant cost a few hundred dollars in an economy where an accountant can be paid that much per hour. Is it any wonder the tax and accounting software industry is booming? In general, the cost of having a human do a physical task is low (health aids, nurses, drivers, machine operators, etc.) and the cost of a robot to do the same task is very high. The cost of having a human do a ‘cognitive task’ is high while the cost of a new computer is very low. Business economics is harsh. Choosing between replacing a cheap worker with and expensive machine versus buying a cheap machine to replace an expensive worker is a no brainer. Who should be worried: the brain worker or the brawn worker?
But what about creativity? There is no machine for that!
In 2013 Researchers Karl Frey and Michael Osborne of Oxford University released a study on the “The Future of Employment: How Susceptible are Jobs to Computerization?” In it they analyzed 700 occupations across 6 dimensions to assess the likelihood that the occupation could be done by a computer or robot. As we might expect, jobs that primarily fall in the “routine” categories of the quadrant chart are near the top of the ‘hit list.’
However, when we look at the range of tasks associated with creative jobs in the Public Relations and Marketing fields, as well as those in the Social Media industry there are some surprises. If you are a marketing manager, they say your job is safe. You are #689 on the list of 700. If you primarily do background market research (#335) or do analysis of statistics (#334) for social media, the researchers claim you have about a 50/50 shot of being replaced by a $200 computer on a shelf. But if you are one of the hundreds of proofreaders or copy editors (#219) or Sales Representatives (#216), you should be a little worried.
You can probably tell that the Oxford predictions suggest that occupations that require people skills are fairly safe. Being able to understand what clients need and what will motivate customers are outside the skill set of today’s machines. Coming up with creative solutions to novel problems is also a challenge. In general, if you find your job to be kind of boring you need to think that there is a computer algorithm with an eye on your paycheck, and those algorithms work 24 hours a day, 365 days a year. They never ask for time off for the kid’s school play, and they don’t have a bad day.
But, we are pretty sure those computers will never have a brilliant creative insight, nor will they come up with the perfect marketing pitch, right? Creativity is a uniquely human skill – the antithesis of ‘robotic, right?’
Here is where things can get a little scary for the creatives among us. You probably remember when IBM’s Watson computer played on the Jeopardy game show. It beat the Jeopardy Champions. The issue was not so much that it won, but how it won. Watson had access to what was effectively its own personal copy of the entire Internet. And, to get the Jeopardy question, it generated literally thousands of possible interpretations of the clue. It then researched every one of those thousands of guesses in less time that the human players could come up with a question.
“The best way to have a good idea is to have a lot of ideas.” This was said by Linus Pauling, one of only four people to be awarded two Nobel prizes. This was a guy who knew about creativity. If computers are going to be a threat in the creative arena (and I am not saying they are) they will be a threat because they can literally run through thousands of possible ideas in seconds and throw out the bad ones, keeping the ideas that are possibly quite good by human standards. Creativity by brute force.
What to do to get ready for the AI Apocalypse
If you are making a living doing a boring, repetitive job, think about what you need to do to boost your creativity. Despite the threat posed by Watson, it will be a while before there will be a Watson on every shelf. If you primarily crunch numbers for conversion rates, SEO optimization, or do background market research – it is getting to be time to retrain. One thing that saved all those workers put out of work by machines in the 1950’s was their ability to adapt, to learn new skills and to stay one step ahead of the machines.
Work with the machines as tools. Let them do the number crunching, the proofreading, the copy editing while you work with the clients. Machines are really lousy at figuring out what the client really wants, what will resonate with the customers. If the machines get good at generating creative ideas – put them to work for you. Let them work all night to come up with some candidate campaigns. Then you can come in and assess which ones are the best, and take the credit. They are just machines after all.
About Louise and Jim Gunderson: They have been working on Artificial Intelligence (AI) and robotics for a combined 30 years. With PhDs in Systems Engineering (Louise) and Computer Science (Jim), the Drs Gunderson have a ideal perspective on both the technology and the impacts of the the AI revolution. They are the authors of numerous technical papers, the book “Robots, Reasoning, and Reification.” In addition to extensive writing, they have done numerous national and international presentations on topics ranging from the social effects of AI to a TEDx presentation on consciousness and robotics.
Beyond the technical side they are serial entrepreneurs focusing on applying AI to improve business processes and are currently working on CAIT – an AI system that does content generation for Social Media. Longtime residents of Colorado, they share a small Denver Bungalow with their Australian Cattle Dog “Sothis”, and can occasionally be found chasing storms across Tornado Alley.