Preparing my child for unemployment
Part 1 — The threat of robots
As a father, I have been described as anxiety-ridden. I wish I can just sit back and sip my coffee while my boy eats his ice-cream, sprinkles it with sand, licks it, drops it, cries about it, picks it up, and licks it again. But I cannot. Unsurprisingly, I worry about what my 3 year old will do when he is 30.
Predicting the future
I am not oblivious to the folly of predicting the future, but:
- I know people can make themselves less terrible at it, and
- specially when looking at general trends (e.g climate, not weather).
Most importantly, I need predictions (accurate or otherwise) about the future, to make decisions about the present. For example, if I think it will rain later, I would bring an umbrella, even when I know I could be wrong. So before I make a decision on my son’s education, first I need some predictions (assumptions/working hypotheses) about the future of work.
Threat of robots
There are plenty of articles on “jobs most/least likely to be automated” floating around. Some even claim that “by 2045, computers may be able to do almost any human task, and unemployment will sit close to 50 percent”. So you’ve probably read something about robots taking over, with dramatic terminator references. You can downplay this threat in 2 ways:
- It won’t happen to me: While many of us (65%) believe robots can displace human jobs, most (80%) don’t expect this to happen to them. This doesn’t really deserve further discussion…
- Luddite fallacy: Conventional (economic) wisdom tells us that when technology replaces human workers, it creates more jobs in other ways. For example, during the Industrial Revolution, “automation in textiles, steel-making, and a whole range of other industries led to a major increase in manufacturing jobs”. The recent threats about automation are just “the latest chapter in a long story called The Boys Who Cried Robot”.
Suck at everything
Unlike previous disruptions such as when farming machinery displaced farm workers but created factory jobs making the machines, robotics and AI are different. Due to their versatility and growing capabilities, not just a few economic sectors will be affected, but whole swaths will be.
So even if technology can create more jobs, the question is whether human are better at these new jobs? For example, web developer is a relatively new role created by the rise of internet, but the demand of web developers (search trends below) has been dropping steadily since 2006. In contrast, computer assisted self-service solutions like Squarespace and Wix shot up. Furthermore, this is not even the endgame for automation — services like the Grid are taking it further by creating “AI websites that design themselves”.
Perhaps you side with the 52% of experts who disagreed that by 2025, networked, automated, artificial intelligence applications and robotic devices will displace more jobs than they create. In this instance, I have no trouble betting on the 48% minority who do foresee mass displacement, specially given that some of the n0-displacement group only disagreed with the timeframe of 2025 (my son will still be in high school).
The threat of mass displacement doesn’t even require artificial general intelligence (one that could perform any intellectual task of a human). Consider the most-common occupations like retail sales, cashier, food and beverage server, and office clerk (10% of the US labor force) — all of these jobs could be automated, with existing artificial technology. Even when AI cannot take over your role completely, it can still take your job by enabling substitute services. For example, while Uber cannot provide self-driving cars right now, their ride matching/pricing system can still displace a bunch of taxi drivers (probably more than the 6,700 Uber employees…).
This is not just a problem for “lower-end” work. Even doctors are on the chopping block now, because computers could eventually replace 80% of what doctors do — with checkups, testing, diagnosis, and prescription all done by sensors, passive and active data collection, and analytics. These days, computers not only can beat us with brute force (process more data), it can also defeat us in traditional “human” cognitive abilities like heuristics and pattern recognition, evident in epic battles like Deep Blue vs Kasparov and AlphaGo vs Lee Sedol.
This threat is not just my imagination, it’s what a whole bunch of AI experts expect to see. Stanford has a 100-year effort to study and anticipate the effects of AI, and their 2016 report says AI will displace cognitive jobs like the way it did with factory jobs. It predicts that a large fraction of the total workforce may, in the long run, lose well-paying “cognitive” jobs.
As labor becomes a less important factor in production as compared to owning intellectual capital, a majority of citizens may find the value of their labor insufficient to pay for a socially acceptable standard of living.
The human touch
Once you accept that computers can and will outperform us cognitively, the most common rebuff is that we still need/demand the “human touch”, like:
- a pimply boy to tell us about the fish of the day,
- a bank teller’s friendly smile instead of the cold hard ATM, and
- a doctor who loves telling people that they only have 6 weeks to live.
In case you haven’t picked up on my sarcasm, there is simply no evidence that “human touch” cannot be replaced for many areas of our lives. Instead of taking you on a giant leap to a world of robo-hubbies, let us ease into it with a slippery slope:
- We always enjoyed non-human companionship, like pets. Nowadays, they are not just a supplement to human interactions, but a replacement in some cases. For example, some women are now choosing dogs over motherhood, saying that they would “rather have a dog over a kid”.
- For most of our human interactions, the goal is not the interaction per se. We want food to eat, transportation to get around, clothes to wear…etc, and the interaction with the goods/services provider is just a byproduct. From vinyl records to vending machines, we have been cutting out the “middleman” for a long time, even without the help of AI.
- In many aspects, having goods/services without the intermediary person is not something we endure, but what we actually prefer. For example, we created the front-facing camera (and selfie sticks), to replace the stranger who might run off with our phone/camera. And now that we’re free from this risky transactional human interaction, we can get on with building even better machines — from face-detecting cameras to a drone that follows you around like your own paparazzi.
- Surely there are some interactions that are so intimate and “human”, they cannot be replaced by machines, like…sex? Nope, non-human objects have been used to satisfy that need for a long time, so the rise of sexbots is a “no-brainer”. As noted by Dr. Henrik Christensen, chairman of the European Robotics Network, “people are willing to have sex with inflatable dolls, so initially anything that moves will be an improvement.”
- What about the your children? Surely we won’t let robots look after our sweet little munchkins. They need love and care, which can only be provided by poorly paid teachers or teenage girls. Seriously, if you cannot imagine leaving kids with interactive machines designed to be fun and educational, just look for a kid with an iPad — they’re not hard to find.
Coming Up Next…
Unlike Sarah Connor, I am not preparing my son for robots hunting him down for sport. My fear is much less imaginative. Even with just half of the predicted 50% unemployment, we’d be on par with the Great Depression.
While I am pretty certain about the “mass” displacement of human labour by machines, I am less certain about how we will respond to it as a society. Universal Basic Income is probably the most prominent response to the automation threat. So that’s the topic I will examine next, to see how I can best prepare my child for a life of unemployment.
If you liked this, click the💚 below so other people will see this here on Medium.