Listen to the audio version of this article!
When it comes to the impact of Artificial Intelligence (AI) and Automation, there is no debate that advances in these areas will engender profound changes in our world. Rather, the debate centers on what these changes might look like.
There are many who express concern or even outright fear about the impact of artificial intelligence on our future, and with good reason. A recent report from Forresterpredicts that by 2021, intelligent agents and related robots will have eliminated a net 6% of jobs. That’s a huge number: approximately 9 million jobs in the United States.
And that’s just by 2021. A widely noted study, “The Future of Employment: How susceptible are jobs to computerisation?”, estimates that 47% of all US jobs are at risk.
“Stop being a Luddite”
The Luddites were textiles workers who protested automation, eventually attacking and burning factories because, “they feared that unskilled machine operators were robbing them of their livelihood”. The Luddite movement occurred all the way back in 1811, so concerns about job losses or job displacements due to automation are far from new.
When fear or concern is raised about the potential impact of artificial intelligence and automation on our workforce, a common response is thus to point to the past; the same concerns are raised time and again and prove unfounded.
In 1961, President Kennedy said, “the major challenge of the sixties is to maintain full employment at a time when automation is replacing men.” In the 1980s, the advent of personal computers spurred “computerphobia” with many fearing they’d be replaced by computers.
So what happened?
Despite these fears and concerns, every technological shift has ended up creating more jobs than were destroyed. When particular tasks are automated, becoming cheaper and faster, you need more human workers to do the other tasks in the process that haven’t been automated.
“During the Industrial Revolution more and more tasks in the weaving process were automated, prompting workers to focus on the things machines could not do, such as operating a machine, and then tending multiple machines to keep them running smoothly. This caused output to grow explosively. In America during the 19th century the amount of coarse cloth a single weaver could produce in an hour increased by a factor of 50, and the amount of labour required per yard of cloth fell by 98%. This made cloth cheaper and increased demand for it, which in turn created more jobs for weavers: their numbers quadrupled between 1830 and 1900. In other words, technology gradually changed the nature of the weaver’s job, and the skills required to do it, rather than replacing it altogether.” — The Economist, Automation and Anxiety
Impact of Artificial Intelligence – A Bright Future?
Looking back on history, it seems reasonable to conclude that fears and concerns regarding AI and automation are understandable but ultimately unwarranted. Technological change may eliminate certain jobs, but it has always created more in the process.
Beyond net job creation, there are other reasons to be optimistic about the impact of AI and automation.
“Simply put, jobs that robots can replace are not good jobs in the first place. As humans, we climb up the rungs of drudgery — physically tasking or mind-numbing jobs — to jobs that use what got us to the top of the food chain, our brains.” — The Wall Street Journal, The Robots Are Coming. Welcome Them.
By eliminating the tedium, AI and automation can free us to pursue careers that give us a greater sense of meaning and well-being. Careers that challenge us, instill a sense of progress, give us autonomy, and make us feel like we belong; all research-backed attributes of a satisfying job.
And at a higher level, AI and automation will also help to eliminate disease and world poverty. Already, AI is driving great advances in medicine and healthcare with better disease prevention, higher accuracy diagnosis, and more effective treatment and cures. When it comes to eliminating world poverty, one of the biggest barriers is identifying where help is needed most. By applying AI analysis to data from satellite images, this barrier can be surmounted, focusing help in the most effectively.
Impact of Artificial Intelligence – A Dark Future.
I am all for optimism. But as much as I’d like to believe all of the above, this bright outlook on the future relies on shaky premises. Namely:
- The past is an accurate predictor of the future.
- We can weather the painful transition.
- There are some jobs that only humans can do.
The past isn’t an accurate predictor of the future
As explored earlier, a common response to fears and concerns over the impact of artificial intelligence and automation is to point to the past. However, this approach only works if the future behaves in similar way. There are many things that are different now than in the past, and these factors give us good reason to believe that the future will play out differently.
In the past, technological disruption of one industry didn’t necessarily mean the disruption of another. Let’s take car manufacturing as an example; a robot in automobile manufacturing can drive big gains in productivity and efficiency, but that same robot would be useless trying to manufacture anything other than a car. The underlying technology of the robot might be adapted, but at best it still only addresses manufacturing
AI is different because it can be applied across a broad range of industries. When you develop AI that can understand language, recognize patterns, and problem solve, disruption isn’t quarantined. Imagine creating an AI that can diagnose disease and handle medications, address lawsuits, and write articles like this one. Oh wait, you don’t have to imagine, AI is already doing those things.
Another important distinction between now and the past is the speed of technological progress. Technological progress doesn’t advance linearly, it advances exponentially. Just take Moore’s Law, which has observed that the number of transistors on an integrated circuit doubles every ~2 years.
What do you get when technological progress is accelerating and AI can do jobs across a range of industries? An accelerating pace of job destruction.
“There’s no economic law that says ‘You will always create enough jobs or the balance will always be even’, it’s possible for a technology to dramatically favour one group and to hurt another group, and the net of that might be that you have fewer jobs” —Erik Brynjolfsson, Director of the MIT Initiative on the Digital Economy
In the past, yes, more jobs were created than were destroyed by technology. Workers were able to reskill and move laterally into other industries instead. But the past isn’t always an accurate predictor of the future, we can’t complacently sit back and think that everything is going to be ok.
Which brings us to another critical issue…
There will be an extremely painful transition
For a second, let’s pretend that the past actually will be a good predictor of the future; jobs will be eliminated but more jobs will be created to replace them. This brings up an absolutely critical question, what kinds of jobs are being created and what kinds of jobs are being destroyed?
“Low- and high-skilled jobs have so far been less vulnerable to automation. The low-skilled jobs categories that are considered to have the best prospects over the next decade — including food service, janitorial work, gardening, home health, childcare, and security — are generally physical jobs, and require face-to-face interaction. At some point robots will be able to fulfill these roles, but there’s little incentive to roboticize these tasks at the moment, as there’s a large supply of humans who are willing to do them for low wages.” — Slate, Will robots steal your job?
Blue collar and white collar jobs will be eliminated, basically anything that requires middle-skills (meaning that it requires some training, but not much). This leaves low-skill jobs, as described above, and high-skill jobs which require high levels of training and education.
There will assuredly be an increasing number of jobs related to programming, robotics, engineering, etc.. After all, these skills will be needed to improve and maintain the AI and automation being used around us.
But will the people who lost their middle-skilled jobs be able to move into these high-skill roles instead? Certainly not without significant training and education. What about moving into low-skill jobs? Well the number of these jobs are unlikely to increase, especially as the middle-class loses jobs and stops spending money on food service, gardening, home health, etc.
This is why the transition could be extremely painful. It’s no secret that rising unemployment has a negative impact on society; less volunteerism, higher crime, and drug abuse are all correlated. A period of high unemployment, in which tens of millions of people are incapable of getting a job because they simply don’t have the necessary skills, will be our reality if we don’t adequately prepare.
So how do we prepare? At minimum, by overhauling our entire education system and providing means for people to re-skill.
To transition from 90% of the American population farming to just 2%during the first industrial revolution, it took the mass introduction of primary education to equip people with the necessary skills to work. The problem is that we’re still using an education system that is geared for the industrial age. The three Rs (reading, writing, arithmetic) were once the important skills to learn to succeed in the workforce. Now, those are the skills quickly being overtaken by AI.
For a fascinating look at our current education system and its glaring faults, check out this video from Sir Ken Robinson:
In addition to overhauling out entire education system, we also need to accept that learning doesn’t end with formal schooling. The accelerating pace of technological change means that learning must be a lifelong pursuit, constantly re-skilling to meet an ever-changing world.
Making huge changes to our education system, providing means for people to re-skill, and encouraging lifelong learning can help mitigate the pain of the transition, but even that isn’t enough…
99% of jobs will be eliminated
This claim may seem bold, but it’s all but certain. All you need are two premises:
- We will continue making progress in building more intelligent machines.
- Human intelligence arises from physical processes.
The first premise shouldn’t be at all controversial. The only reason to think that we would permanently stop progress, of any kind, is some extinction-level event that wipes out humanity. Otherwise, progress will continue to be made, the incentives for people/companies/governments are too high to think anything different. And it doesn’t matter how fast that progress is, all that matters is that it will continue.
The second premise will be controversial, but notice that I said human intelligence. I didn’t say “consciousness” or “what it means to be human”. That human intelligence arises from physical processes seems easy to demonstrate: if we affect the physical processes of the brain we can observe clear changes in intelligence. Though a morbid example, it’s clear that poking holes in a person’s brain results in changes to their intelligence. A well placed poke in someone’s Broca’s area and voila, that person can’t process speech.
With theses two premises in hand, we can conclude the following: we will build machines that have human-level intelligence and higher. It’s inevitable.
We already know that machines are better than humans at physical task, they can move faster, more precisely, and lift greater weight. When these machines are also as intelligent as us, there will be almost nothing they can’t do. So 99% of jobs will be eliminated.
We will still need leaders (unless we give ourselves over to robot overlords) and the arts/music/etc. may remain solely human pursuits too, but just about everything else? Machines will do it, and do it better.
“But who’s going to maintain the machines?” The machines.
“But who’s going to improve the machines?” The machines.
- It’s unlikely that the future will play out like the past. There’s no guarantee that more jobs will be created than are destroyed by AI and automation.
- Even if the future does play out like the past, the jobs being created will require re-skilling and better education. These services aren’t currently provided, so unless we make major changes, we’ll have millions of people who can’t get jobs and we’ll all suffer from it.
- Even if we manage to deal with this transition effectively, all jobs will eventually be eliminated by machines.