We are switching things up!

This publication is currently undergoing construction and we will be re-launching soon!

Keep an eye on our Instagram and Facebook for updates.

The Employment Terminator


Are robots coming for your job? Maybe, writes Brendan Byron.

I bet I’d surprise no one if I told you that jobs are harder to find these days. Every “entry level” position wants 3-5 years of industry experience, and even a postgraduate degree is no guarantee of stable employment.

What might surprise you, though, is this: there’s good reason to believe this lack of jobs is about to get much, much worse over the next twenty years. And I’m not blaming immigrants or austerity or recessions, either.

By far the biggest challenge to your employment prospects is the rise of the robots.

It sounds bizarre, but if you think about it, this really shouldn’t come as a shock – tellers in banks were replaced by ATMs a decade ago. In the last five years, supermarket checkouts that used to involve a dozen workers in a dozen aisles have been rapidly replaced by one worker supervising a dozen checkout machines. In the last six months, ticket windows at train stations have all but closed, with thousands of workers laid off.

Which, don’t get me wrong here, is great news. It’s creative destruction – technological innovation kills old jobs like abacus makers and all the people who would have made abacuses turn to improving computers instead. All the taxi drivers and truckies put out of work by driverless Google cars will find new jobs in the next great endeavours – jobs that haven’t even been invented yet.

Or will they?

Last week I sat down with UNSW Senior Economics Lecturer Stanley Cho. He’s an expert in ‘capital-skill complementarity’ – the tendency of skilled individuals to earn more and be in more demand as capital, or machines, become more widely-used. And while he stuck to the standard script most of the time – that new jobs will always be invented – he did say there was evidence to the contrary.

I asked him: if Google’s driverless cars put a million people out of work (a not-unreasonable assumption), will a million jobs be created in turn? He said “Maybe not.” Dr Cho referenced America during the Global Financial Crisis: the economy grew for years before the job market began to recover as well. Dr Cho theorised that to cut costs, companies replaced people with cheaper capital.

“All of these unskilled workers were replaced with cheap machines, secured with low interest rates and cheap loans,” he said.

You might think you’re safe with your readily applicable, highly prized Comm/Law degree. You aren’t. Going into finance? Computers trading stocks with computers already comprise over 99% of financial transactions – sometimes with devastating consequences.

On May 6, 2010, at 2:42pm, the Dow Jones plummeted out of the blue. Trading algorithms got caught in a loop, buying and selling the same stocks over and over. Within ten minutes of the malfunction, a trillion dollars disappeared from the global economy – 9% of the Dow Jones index’s total value. Twenty minutes later, at 3:07, the marketbots had self-corrected. Not only are humans being outmatched by machines, when it comes to things like high-frequency trading, we’re not even playing the same game anymore.

And here’s where we get down to the crux of it: the reason this time’s different – and why I doubt today’s taxi drivers and financiers won’t find new, better-paying jobs when their old ones get mechanised – is because there’s something fundamentally different in how this generation of technology plays out. We’re not building machines with better strength or better speed than humans — we’re building machines with better minds than humans. And some time in the future – probably not in the next 20 years, but not “never” either – we’re quite likely to crack the code and build artificial intelligence that can think creatively better than a person can.

That’s a big deal. At that stage, you’re not just worried about the employment prospects of people when computers can do any job better. Because the next job to go might be computer programmers. Why couldn’t an AI learn to improve its own cognitive ability? If it improves its own cognitive ability, what stops it from getting even better? Where is the roof?

Elon Musk, the genius behind PayPal, Tesla Motors, and SpaceX, goes even further. He wrote on twitter in August last year: “Hope we’re not just the biological boot loader for digital superintelligence. Unfortunately, that is increasingly probable”.

These are a lot of big ideas. I don’t ask you to follow me all the way from mechanised cars to digital superintelligences – you can call me crazy at some point along the chain, and you probably should. It’s not clear that any of these potentialities might happen. What is clear, though, is that some tiny, miniscule fraction of it will.

And we better be prepared for when it does.