The Short Answer: Probably Not. Donald Farmer, Qlik’s VP of Innovation and Design, believes the current automation trend will take longer than we think.
The Industrial Revolution was responsible for changing the job landscape in the 1800s. At the time, the world was worried that the new technology would eliminate a majority of the need for human workers. Instead it created millions of jobs: enough to employ the 20th century’s booming population.
The job market of today is approaching another crossroads in regards to technology and employment and people are finding themselves with questions: will I still have a job once the robots move in? Two hundred years ago, 70 percent of American workers worked in agriculture; today, only one percent do. Will automation in the modern-day workplace lead to this same trend?
It’s true that automation continues to be a growing trend, but not everyone is pessimistic about the job market outcome. Donald Farmer, VP of innovation and design at Qlik, is part of the “glass-half-full” camp.
“I’m not worried about artificial intelligence,” he said to us in an exclusive interview. “I’m more worried about artificial stupidity.”
With a decade in business intelligence experience with Microsoft, time with startups and consultancies, and a stint as a research historian and an archeologist, Farmer has always been interested in how humans structure knowledge.
“My studies and my career have always led me to think about how we make decisions,” Farmer said. “What technology and its researchers are considering now is the extent to which decisions can be automated. The more we have learned about how human beings understand information and structure a decision, the more we learn just how necessary humans are for decision making.”
Twenty years ago, thousands of decisions were complicated. An overdraft on your bank account meant meeting with the bank manager so he or she could scrutinize you on your personal worth and financial viability. Now many, if not all, financial institutions offer overdraft protection, and are immediately able to handle the situation without a human presence. People can even get quotes on home or car loans in seconds online. In short, technology can now make a lot of decisions we used to ask people to make: technology has caught up.
But it took more than a bit of code to get this sense of automation to work: people also needed to trust the system and allow for the deregulation of several systems like those of financial institutions. Now that we trust these systems to protect us when we overdraft or to provide us with an accurate quote for a home loan, what industries does Farmer expect to see deregulating for automation?
“It takes a lot of time for doctors to meet with patients who have a simple malady,” Farmer said. “In the future, we’re looking for machines to be able to assess simple symptoms and assign remedies. I also think petty crimes may be able to be settled out of court by automated systems.”
So if automated systems might be trusted to make these decisions in the near future, what can’t these systems do? As it turns out, quite a bit. Farmer discussed at length the viability of these systems resting on the black-and-white nature of decision making. If the results of the decision can be formed by an algorithm, and are the same time after time—like the symptoms that lead to the diagnosis of a cold or the flu—a machine or automated system can make the decision. If there is any degree of ambiguity, or a judgement needs to be made, a machine or system cannot be trusted to make a definitive, appropriate decision.
“Humans beings will always have the competitive edge,” Farmer commented. “Most decisions are laced with an amount of ‘gray area’. You won’t hire someone because a machine or a program tells you to, you’re likely to base your decision on how that person interacted with you during an interview. Same with strategic decisions or tactical decisions. Automated systems can only provide so much.”
It’s possible—and probable, Farmer said—that as more automated systems move into the repetitive jobs that machines can handle, jobs could be taken away from humans in the short term. And while this is a scary pronouncement for some workers, it’s a win in the long term.
Instead of doing trivial jobs that don’t offer challenges to employees, people will either have more free time to do things that are more interesting to them, or be able to train in a career that truly piques their interests. There are downsides, however, as there are with anything, but Farmer is optimistic that adjustments made in the short term will make the future workforce more satisfied with their jobs.
Farmer isn’t without concerns, however. Although people like Stephen Hawking and Elon Musk have warned against the potential dangers of artificial intelligence, Farmer shared that there was something much more dangerous: artificial stupidity.
“Machines are good at input and output, but once a complication is introduced, the system doesn’t work as it should anymore. If something causes your GPS to give you a direction to turn into a lake, human decision overrides the absurdity of this—for the most part. In my travels, I’m glad I have a highly technical automated system controlling the plane, but I’m even happier that there are two people overseeing that technology.”
According to Farmer, there are two things currently missing from the progress of technology that is slowing down the march towards the dreaded complete automation. For one, machine intelligence is not progressing as fast as many people think: we’re seeing a lot of simple decisions happen quickly, but there is nothing yet that has allowed a machine to handle ambiguity.
And although this may sound like obvious, people tend to forget that machines cannot control themselves.
“If we don’t like a system we are creating, if we think we’re giving it too much power, then we don’t have to continue with it,” Farmer said.
So if we don’t want robots or automated systems to be this century’s receptionists or managers, then they don’t have to be. Technology is moving pretty fast, but not so fast that you’re going to miss where it’s going next.