Machine Learning (a sub-category of AI) is the ability to process huge volumes of data and complete tasks more efficiently than a human. Moreover, ML actually learns. Once you set up a Machine Learning model, it can understand and store different outcomes based upon experiences or other input.
AI Across Industries
Let’s look at an AI process that picks stocks. On the human side, a hedge fund analyst can look at all of the available data, such as management performance, past stock prices, and an outlook for the future, and turn that input along with experience in picking stocks that hopefully go up.
Now, let’s say an application that picks stocks is bound to an ML engine and can consider the same data as the analyst, and also incorporate the knowledge and experiences of 1,000 other financial experts. Using cloud-based processing that’s almost free, the hedge fund administrators can employ as many of these processes as they need. Information is processed 1,000,000 times faster, and stocks are picked and plotted in seconds. That hedge fund analyst who makes $1,000,000 a year with bonuses is easily replaced with a cloud instance that costs $1,000 a month.
Other examples are more blue collar. Self-driving technology that manufacturers have spent years to perfect is now ready for prime time. While this means that we can crawl into our cars and perhaps take a nap on the way home, it also means that self-driving trucks and taxis also become the norm.
According to a report by Goldman Sachs, the effect is pretty devastating. Self-driving cars and trucks could drain 25,000 jobs per month—or 300,000 per year—from America’s bus, taxi, and truck industries when they finally hit the market.
In 2014, approximately 4 million people drove professionally in the United States, 3.1 million of which were part of the trucking industry. That majority, coupled with the fact that companies like Uber and Tesla are already working to develop self-driving freight trans-port technology, means that truck drivers could bear the brunt of the predicted job losses. Additionally, the long distances and demanding nature of the work also make trucking a prime target for automation.
A bit concerning is that those driving trucks and taxis (or Ubers) are likely to be the head of the household, thus the primary breadwinner. As a result, many families could be forced into welfare and food stamp roles until those workers who are displaced by self-driving technology are able to pivot to another career path, which could take more than a year.
Others affected by automation, and specially AI-based automation, include factory workers who have been replaced over the years by robotic technology. While this is nothing new to factories, even the limited number of humans needed to run the robots could find that they are now facing job loss as factories move from 2.0 to 3.0 in robotic technology that actually eliminates humans from machine maintenance jobs. Indeed, AI may enable the “dark factory” in just a few years, where a few people who man a control booth are all that’s needed for a vast production operation.
If you think that those who man data centers, including cloud computing data centers, have safe jobs, you would be wrong. While these impressive four-story windowless buildings, as big as shopping centers, would make you think of thousands employed inside, they rarely have more than a couple of dozen people who repair the physical equipment.
The World Economic Forum expects automation and Artificial Intelligence (AI), to result in the loss of at least 5 million jobs globally by 2020.
The data centers are monitored consistently by automated and intelligent engines. Only once or twice a month do actual humans enter the data center to replace failed servers and other equipment. Older data centers that are more manual in terms of leveraging humans are considered cost inefficient and are being replaced by increasingly automated data centers, which typically leverage AI and other technologies.
Point the Compass in a New Direction
So, what’s a human to do in a world that’s becoming more populated with robots, and even robots that can do your job? It’s a matter of pivoting to other career directions. Don’t worry, we’ve done this many times before throughout history.
Consider the arrival of the internal combustion engine, and the ability to move quickly across the land without the aid of horses. Those who served the horse and carriage industry, including those who made carriages and buggy whips, had to retool and find other ways to stay relevant. Those who worked for those industries had to figure out new career paths.
We have gone through many iterations of this shift in the last hundred years, including the rise of computers, the rise of instant communications, the rise of online stores… And the list goes on. These days, it’s a matter of looking at what parts of the job market will grow, or are growing.
The number of job postings is increasing in the machine learning space, which is not surprising. However, these are likely to be taken by those already on a computer career track, and thus not helpful to displaced truck drivers. However, it is a good option for those who are out of data center jobs or other areas of high technology that are affected by the growth of automation and AI.
The dilemma is that AI, and automation in general, will take on much of the heavy lifting that’s currently done by humans. This shift should make things safer and cost less when you consider the impact on most lives. However, it will also take more jobs than it creates, and those who have been replaced by AI-based automation and robotics won’t have the same options as those in the high tech industries, who already have to reinvent themselves about every 10 years to survive.
Transitioning to an AI World
In a recent interview with Quartz, Bill Gates said that a robot tax may be the way to go to allow for a better transition to automation and AI-based job killing technology. “These taxes could finance jobs taking care of elderly people or working with kids in schools, for which needs are unmet and to which humans are particularly well suited.”
Gates argues that governments must oversee such programs rather than rely on businesses to do so since they are profit motivated. The government can redirect the jobs to help people with lower incomes, or the displaced blue-collar workers discussed above. Indeed, EU lawmakers considered a proposal to tax robot owners to pay for training workers who lose their jobs. However, in February of this year, legislators ultimately rejected it.
There are no easy answers here. These are old problems in a modern setting, in terms of job killing automation. But the new AI technology evolution could be distinctly different from prior evolutions. Why? AI is smarter than we are.