Tongues have been wagging about the rise of the robots. Robots taking over manufacturing, robots acting as home companions and robots taking our jobs. A lot of digital ink has been spilled on the subject. You know what? Just about every one of these fear mongering articles… is right! Many of the most common types of robots won’t look anything like people. Don’t expect tall silvery humanoids with antennae sticking out of their heads. Many robots will be hard to recognize or they might just be one more app running on a computer, or even a cell phone.
What makes robots and artificial intelligences game changers is not how they look, but what they do. And what they do… or are about to do… is your job! When robots aren’t actually performing your job, they will spend their leisure hours learning how to do your job better. Or, maybe they will learn how to do your boss’s job. Before the end of 2016 you’re going to see some of the first skirmishes in the robot revolution. By 2017 you’ll see human employees lose a few critical battles. Read along and learn what you need to know about the coming Robot Revolution!
Machines replacing people? This isn’t exactly new. Machines, starting with simple machines like the wheel, have made people more productive. For most of human history, humanity was so poor that we would immediately consume any additional productivity. Especially when that productivity was in the form of food or fuel. This additional productivity led to more wealth and more security. This Virtuous Cycle has been working since the earliest days of civilization.
Virtuous Cycle: The age of Sail started thousands of years ago and ended just a few centuries ago. Sail driven ships moved more goods, for a lower cost. It also showed us that there were far more goods in the world than our own village, and some foreign goods (high-quality wool, silk cloth, spices, delicate jewelry, superior metals) were vastly superior. We had choices in what we could buy. This drove wealth, then luxury, and finally leisure time. Better machines improved productivity, which improves our standard of living. There were gaps and exceptions, but the story of mankind has been one of growing wealth.
Political collapse, war and natural disaster could disrupt this Virtuous Cycle, but eventually it would move forward and increase our wealth. In the 21st century we live with constant technological change and constant material improvement. There may be a good argument that wages have stalled for the last 20 years, but even with no more dollars in our pockets, our lifespan increases every year, diseases are cured, nutrition around the world gets better, and even our entertainment choices improve. But, what if something broke this cycle? Something like… robots?
Consider how this “Virtuous Cycle” works. It takes a very long time to implement new technology, and even longer to fully optimize and integrate into society. Social and technological limitations (cost of travel, cost of fuel, short live spans) made it difficult for the technology to spread very far from where it was discovered. It took centuries for the locals to figure out all of the details of a new technology, and even after a thousand years, it might not spread very far. Over this long period of time, wars and disasters could disrupt or even wipe out the technology. But all of these problems provided people and society
But all of these problems provided a benefit. People and society had time to absorb the technology. 3,000 years ago iron was the “secret weapon” the Hittite people used to conquer the middle east. In less than a thousand years, the iron age had spread across most of the world and even small villages had blacksmiths.
American Example: If an iron age example a bit too “dawn of time” for you, let’s move it up to… the early United States. America started out as a largely agricultural country, but also had fishing, mining, a textile industry, fur trapping, shipbuilding and other industries. By 1850, agriculture was America’s the primary source of employment. The industrial revolution arrived and replaced animal and human labor with machines. EXCEPT the American South. The South’s choice of slavery over technology sparked a war between the North and South. Over 150 years, farm workers declined from 70% of the workforce to just 2%.
Need another example? Here’s one! The factories that built the machines that replaced the farmers, needed workers. Early factories were often difficult places to work, but they were better than farms. They paid more, they brought people together, increased education, and (eventually) improved life expectancy. Were the farmers happy about the transition? Absolutely not. There were all political and physical fights over land and jobs, culminating in the Civil War. Nonetheless, the industrial age turned farmers into factory workers, and by 1960 most Americans worked in factories. Yet just 50 years later, manufacturing is just 8% of the job market, and shrinking.
There is, however, an important difference between the revolution in farming and the revolution in farming and the revolution in manufacturing. Time! You may not be surprised that the Iron Age took 1,000 years to take hold. America’s farming revolution took just 150 years. Manufacturing took just 50. Every new technology revolution builds on the last revolution and takes less time to be adopted across a larger area. Still not enough? Let’s look at one more example.
Chinese Comparison: In 1980, 70% of China’s workforce was agricultural… just like the U.S. in 1850. China needed to move agricultural workers into manufacturing jobs. This became the #1 priority of their long-term economic plan. By 2012, agricultural workers were just 35% of the workforce. In percentages, that’s at least twice as fast as the US conversion. Pretty good. But in total population, this is absolutely stunning!
America eliminated most of the agricultural positions when our nation was young, with a small population. Agricultural employment peaked in the US in the 1920’s, at just under 11 farmers; now there are just 2-3 million farmers. This transformation occurred over 7-8 generations, so it impacted the lives of more than 11 million farmers. However, China peaked at 700-800 million farmers towards the end of the 20th century. China eliminated 350-400 million jobs in just 35 years, or 10 million annually. Rather than eliminating jobs twice as fast as America, it might be more accurate to say that China eliminated jobs 100 to 200 times faster. How long would it take China to eliminate the 40 to 60 million corporate “knowledge worker” jobs in America?
New Cycle: The Data Economy arrived in the last quarter of the 21st Century. Instead of wealth arising from the land or a factory, it came from information. Farms were largely self-supporting through human history. In the 20th century, farms grew large enough to need significant capital to buy fertilizers, pesticides, seeds and equipment. Manufacturing firms were still larger, and needed banks for funding, banking services, and services to exchange payments with workers, vendors and customers. Financial firms (and financial departments in corporations) needed Knowledge Workers.
Knowledge workers turn collections of information into knowledge (insights). These insights guide investments: which markets are growing, where are profits the highest, etc. Knowledge workers initially used data, a calculator and paper & pen. Soon, computers replaced calculators, and data feeds replaced paper. Demand continued to rise, and publishing firms sold on-line data: McGraw Hill Financial, Standard and Poors, Moody’s, Gartner and others. On-line services began replacing jobs. Computer and consulting firms (IBM, Oracle, Accenture, McKinsey, Bain, and scores of others) sold services to automate (or outsource) functions, reducing onshore knowledge worker jobs.
Outsourcing created the Virtuous Cycle 2.0, and a quandary. Jobs with better salaries were being created but in lower pay geographies. A $60,000 position in the US, had a total cost of $100,000-$120,000 (adding bonus, benefits, space, IT support, etc.). In India, that position is paid just $20,000 (the employee is paid just $6,000). Even if it takes 2-3 workers to replace one onshore, it still saves money. That $6,000 offshore salary isn’t much in the US, but it an be a great salary offshore. Virtuous Cycle 2.0 still works, if you look at global results and ignore domestic losses. Did VC 2.0 continue to create wealth? This is still being hotly debated.
The Vicious Cycle: The Data revolution differed from the two prior revolutions in many ways, but the key difference was… speed! It’s very simple. Employment is determined by new jobs created and old jobs lost. Whether the economy is hot or cold, jobs are always created and lost.
In the Data economy, computer intensive work was targeted for offshoring. In 1970, computer jobs were a rarity. By 2000, you couldn’t get hired in a corporation without computer skills. One by one, isolated parts of the corporation that were not “outsourcing ready”, disappeared. Decades of effort, effort that was not always specifically aimed at outsourcing, slowly prepared millions of jobs for outsourcing and automation. “Flexible” robots and A.I.s will now trigger the elimination of these jobs.
The Domino Theory: Have you ever watched a domino display fall down? A good display can take days to set up, and seconds to fall. It just takes one little push on that first domino, and the rest all fall. We’ve been setting up “dominos” for the robot revolution for 30 years. Take a good look at this delicate stack of dominos:
- Infrastructure: Think of mobile robots (self-driving cars, flying drones) as blue collar workers, and A.I.s running on computers as white collar workers. Corporations own BILLIONS of PC’s. Each computer could run an A.I. that writes stories (reporter, financial analyst, lawyer), analyzes data (accountant, financial analyst, lawyer) or performs other tasks. Tasks too complex for any single PC would be shared across a network, in your office or around the world. Any work performed on a PC is ready for an A.I. to take it on.
- Capabilities: A.I. systems are disembodied intelligences that read electronic data and documents. However, not all information is in computer format. To capture information from the “real world”, you need cheap and reliable robot “senses”. A generation ago, these didn’t exist. Now they are everywhere, including your cell phone. Robots can read and speak (Siri and Cortina), hear (voice recognition) and see (phone cameras, object recognition software). Touch screens let you physically interact with technology. Smell and taste? Not yet, but specialty sensors can identify: smoke, explosives, drugs, etc. Arms and legs are difficult to mimic, but automated kiosks (such as ATM’s) can take your cash and checks and hand you money without lifting a robotic finger.
- Capacity: Corporate computers are already used by employees, but only for 40 hours a week. “Robot workers” could exploit the remaining 128 hours. As computers wear out or become obsolete, less expensive yet more powerful replacements will lower the cost of robot workers every year. Consider, Deep Blue. In 1985, IBM wanted to build a computer capable of defeating a chess grand master. IBM built the world most powerful supercomputer (Deep Blue), and won! Today, a chess app running on a cell phone can defeat grandmasters. Once a robot takes over a job, they ALWAYS perform it better than a human. A job lost to a robot, is lost forever.
- Self Learning: Robots differ from earlier technology. First, they are flexible, allowing them to work in less structured environments. Older tech would stall when faced with unexpected issues; robots can apply general rules to find a specific solution. That allows robots to work in more environments. Second, robots learn on their own. Humans still write the general software, but robots then work with experts to learn how to do work. The expert gives the robot an assignment. The robot then analyzes documents (spreadsheets, images, forms,etc), to produce its assignment. Just like a student, the robot’s work is graded and the robot eventually “graduates” and can then independently perform work. This process is applied to task after task, eliminating jobs. After “worker” tasks are all replaced, robots can apply the same process to management positions.
- Knowledge Transfer: When outsourcing hit the US, the speed of moving jobs offshore was limited by the speed of training. Training an offshore financial analyst took years. Then, after a few years, the local pool of candidates was smaller or completely used up. More workers could be found by hiring less skilled candidates, and providing more training. This took longer, and cost more, but outsourcing could continue. Once a robot is fully trained, you just download the software to another robot. Or a million. The cost of outsourcing rises every year as the best candidates are consumed. The cost of robots falls, as technology becomes cheaper every year. And each new robot is always more capable than the one before it.
- Scale: In the last revolution, jobs were lost to automation or were moved offshore. In the robot revolution, you will still compete with outsourcers to keep your job. But you will also compete with the robots your company buys, on-line robots from data providers, consulting firms bundling robots and A.I.s with their services, and a new generation of robot vendors. Offshore firms will also build or buy their own robots, and offer hybrid human/robot services. The forces competing for your job will be far more numerous and more capable than ever before.
Robots Everywhere: We’re faced with a pretty simple question, “How many jobs will robots take over and how fast?” The answer is equally simple, “More jobs than ever before, and faster than anything you’ve ever seen!” There are more sources for robots, more corporations and universities working on robots and… thanks to the infrastructure already in place… vastly more jobs for robots to immediately take over.
Robots will create some new jobs, but less than people expect as robots eventually take over their own training, management and possibly sales. In the Virtuous Cycle 3.0, new wealth is still being developed, but fewer are domestic jobs with the same pay. Corporations will have higher profits, retirees will benefit from affordable personal services, and those with wealth will see a new age of luxury.
However, the 30% to 50% of the workforce that will be eliminated, especially the younger workers, will have a more difficult life. Especially in the next 10 years, when we lose 10 – 20 million jobs, or more. In as little as 20 or 30 years, this disruption may end and employment may rise again.
It will be years before we have any idea of what VC 4.0 will look like. Most likely, it will rely heavily on the Creative Economy (more in a soon to be released blog). For now, if you have an entry or mid-level corporate job, spend your time writing or reading reports, work on a computer, drive a car, train, ship or plane, assume that a robot can do your job. If not today, then in a very few years. And that’s my Niccolls worth for today.
Do you agree or disagree? Comment on this blog and tell us what you think!