An extremely belated welcome to article two of Futurism 101! If you missed article one, see Futurism 101: The State of Play.
All image credits: Sebastian Steele
The year is 1900. You are a horse, and you are worried about the future of your job. Those pesky humans are both ingenious and extremely lazy, and over the last two hundred years have been putting steam to good use in all sorts of mechanical muscles. A horse friend, however, reminds you that since all that technology came along your city job has been pretty luxurious – especially compared to old jobs like running into battle and doing farm work. And with so many people moving into the cities, there are bound to be more jobs for horses than ever before. Even when pressed about that new-fangled car-thing, he assures you: “New technology always creates new jobs that we can’t even imagine now”.
Human you, in 2016, knows better. Over time, new technologies arrived which phased out working horses forever. These mechanical muscles could improve much faster than biology could, and just as they pushed horses out of the economy, mechanical minds will do the same to us.
Autonomy is around the corner
The two things which come to mind when you think about “robots” are probably either science-fiction or something akin to a Roomba. What you may not have realised is that robots around the world are getting a lot smarter – a lot more quickly. Remember the whole “exponential growth” thing from last time? Right.
Perhaps one of the most visibly progressing areas is autonomous cars. Many new cars come with some semi-autonomous features, like radar-controlled cruise control, which can match the speed of the car ahead of you. Some cars, such as the Tesla Model S, however, can do a whole lot more. As of a recent software update, the Model S “Autopilot” will let you cruise down a motorway completely hands free. Many auto companies are in on the act, not to mention software giants like Google, and pretty soon we will have cars which drive themselves – not perfectly, but better than humans. Bear in mind that 40,000 people are killed every year in road accidents just in the United States. An autonomous car, meanwhile, won’t get tired, distracted or old.
(Red indicates sensors)
Autonomous cars are already here, with Tesla aiming for complete autonomy in 2018 and Google wanting to commercialise its prototype by 2020. The implications are quite staggering if you consider that the technology behind autonomous cars can be applied to lorries, ships and warehouse robots. All in all, transportation accounts for around three million jobs in the US, and 4.3% of the workforce in the UK – these jobs are over. Consider that where a human lorry driver may only be able to drive eleven hours a day, a driverless lorry could drive nearly the full twenty-four. This technology could effectively double the output of road-based transportation at a quarter of the cost. No technology will replace jobs more quickly than autonomous transportation, and given that typically around a third of a transportation company’s costs are salaries, if you think the unions will save the day, think again. Economics has won in the past, and it will win now.
But this is just the beginning. An oft-quoted Oxford Martin School study from 2013 predicted that forty-seven percent of jobs in the UK have a high probability of being automated by 2033. Let’s take a minute to get to grips with that number: The great depression in the 1930s saw unemployment reach twenty percent in the UK. Twenty percent. The oncoming revolution will see massive changes in how society views work and productivity, and we’re not at all prepared to face it. But maybe you think you’re too highly trained and smart to be replaced by robots? Think again.
But I’m a clever clogs
As University students it can be difficult to remember that your average Joe probably hasn’t been through higher-education, and works in fields which are ripe for automation. Not to single out or belittle those people – as it turns out, many of us are replaceable too. Enter the software bots. These are not programs taught by smart automation engineers, but rather programs which teach themselves skills that a programmer never could. Take the stock market. You may not have realised, but the stock market is largely made up of bots that taught themselves to trade stocks trading stocks with other bots that taught themselves. To be clear, these bots are not following the orders of human controllers – they themselves make the decision to buy and sell.
How about white collar work? Well, a large portion of a lawyer’s work is made up of ‘discovery’, where a lawyer has to sift through thousands of documents to find connections or out of place transactions – this work is already becoming automated in many firms. Likewise, IBM’s Watson supercomputer is currently working as a doctor, understanding people’s own words and returning accurate diagnoses. Watson can be better than any one doctor could be, by learning from other robots around the world to better understand the interaction of each and every drug’s interaction with the body. This is knowledge that one doctor simply cannot hope to learn, and just as it is with autonomous vehicles, a robot doctor needs only be better than a human, not perfect. In the short term, it’s likely that bots like Watson will serve as tools for existing doctors and engineers to advance their respective fields. The kind of software I’m talking about here, however, is only the beginning – Machine Learning, which can be thought of as essentially pattern recognition for computers. The final destination of this type of technology is true Artificial Intelligence – but that’s the story for my next article. What is clear is that even specialised, technical jobs already have room for automation, even with relatively primitive Machine Learning.
But I’m a special creative snowflake
Even more creative jobs such as journalism are already being replaced. Bot-written articles are being published all the time and are usually indistinguishable from human-written ones. Consider the following opening sentences of these two sports articles:
“Things looked bleak for the Angels when they trailed by two runs in the ninth inning, but Los Angeles recovered thanks to a key single from Vladimir Guerrero to pull out a 7-6 victory over the Boston Red Sox at Fenway Park on Sunday.”
“The University of Michigan baseball team used a four-run fifth inning to salvage the final game in its three-game weekend series with Iowa, winning 7-5 on Saturday afternoon (April 24) at the Wilpon Baseball Complex, home of historic Ray Fisher Stadium.”
Could you tell which one was written by a human and which by a bot? Most people can’t, according to a study by Christer Clerwall of Sweden’s Karlstad University. (For those who are curious, the first extract is the bot-written one.) If you ever see articles released within minutes of the news they’re based on, it’s likely to be machine-written. Similarly, the Associated Press uses an automated writing bot to create more than 3000 financial reports per quarter, a practice also carried out by Forbes and more. The co-founder of the Forbes partner Narrative Science, Kristian Hammond, estimated that ninety percent of news could be algorithmically generated by the mid-2020s. Likewise, the management science professor Philip M. Parker has more than 100,000 books for sale on Amazon on topics ranging from Managing Investment Firms to Acne Rosacea. It’s easy to see that with how much “Big Data” is generated these days, it takes software to be able to sift through and analyse it all.
What about automation’s impact on the arts? Well, there are already bots which are capable of creating music, literature and paintings that humans have found pleasing (though perhaps not yet indistinguishable from a human’s work). These bots will only continue to improve and will one day surpass the creativity of a human – all the while being able to produce thousands of works a day. If you think that creativity is some unique, metaphysical aspect of being human, think again – by definition, that which is unmeasurable cannot affect anything in the real world. Just as you have almost certainly read articles written by machines now without even noticing it, so too will you consume other algorithmically-generated media in the near future. People used to think playing chess was a uniquely human skill right until 1997, when Deep Blue beat Garry Kasparov. And as it goes for chess, so it will go for all human talent.
It does indeed look bleak, especially when constantly faced with news such as that of Foxconn (the manufacturer who assembles iPhones and Samsung phones) replacing 60,000 factory workers with robots – just in one factory in China. Similarly, the increased minimum wage of $15 an hour in the US could accelerate the adoption of automation, replacing the jobs of those most in need.
The good news is that certain types of jobs will last longer than others – mainly those with customer-facing positions. Technology is still some way off being able to hold a convincing conversation with a human, and the subtleties of language are likely to evade bots for years to come. That is not to say that these jobs will last forever. Indeed, if there were ever a time for getting ahead while you can and saving for a rainy day, this is that time.
Still, it is clear that there is a tidal wave of change coming in the very near future, and that very few people are aware of it. Thankfully, it is becoming more of a hot topic in the press and perhaps soon will be in mainstream politics too.
You can tell I’m an artist.
What can we do about it?
It is clear from modern history that the forces of economics always win out, so don’t expect the unions to be able to fend off these changes for long. A case in point is ride-sharing services such as Uber, which promise to replace old-fashioned taxis with a more efficient, faster and cheaper service. If you think there is backlash from taxi drivers’ unions now, imagine what will happen when Ubers are autonomous – which, by the way, is Uber’s end goal. But in spite of the backlash and regulation, services like Uber will win in just the same way that Netflix is taking over in the TV space: by offering a better service to the consumer.
So, if you can’t beat them, join them. If we make changes to the way our society works now, we can reap the huge benefits of automation without widening the inequality gap. Measures such as a universal living wage are a good first step, but they need to go much further. The utopian ideal would be, of course, to live in a world of abundance, where no one need work on anything unless they wish to. However, this seems quite far-removed from today’s world and will likely not become a reality for many decades, if at all. A more realistic view is one where there is a reasonable universal wage but people still need to be productive for a living – whether it be via having a customer-facing job, selling hand-crafted products or participating in the emerging digital economies of video games such as DoTA or Team Fortress 2 (a topic worthy of an article in of itself). What is clear is that action needs to be taken now, in preparation for the oncoming revolution rather than in reaction to it. Maybe then we will all be able to reap the benefits of automation and resource abundance, creating a more equal society rather than becoming ever-more divided.