Future of Employment
Jobs in the Age of AI
People from all walks of life are concerned about their futures in the job market.
But the rise of AI is not the same thing as digitalization.
The wave of digitalization, which swept the world in the 1990s and 2000s, allows me to send this article to TBY’s editorial department instead of having to hand it in personally.
AI, on the other hand, would be a bot capable of figuring out what is worth writing about, then writing it with a degree of wit and penmanship that would put William Shakespeare to shame, and releasing it whenever it saw fit.
Such a level of sophistication in AI solutions would free up a lot of my time. The bad news is that I would not even have the chance to vent my bitterness by writing pseudo-literary works of fiction because—and this is very important—if there was a market or demand for that, AI would outperform me in doing it.
But are these predictions going to far? Surely writing and language are the domain of humans and completely out of the reach of computer programs.
Programs powered with computerized linguistics, however, are already well on their way, dabbling with uncanny results in human language.
Think of the task of assigning scores to essays written by students based on criteria such as formal accuracy, choice of words, creativity, formulation, and power of reasoning. We would surely all agree that such a task falls under the domain of humans, wouldn’t we?
The news is that providers of renowned language proficiency tests such as the New Jersey-based Educational Testing Service (ETS)—the organization behind tests like TOEFL iBT and GRE—and the London-based Pearson Education currently rely on artificial intelligence for rating essays and recorded responses to speaking prompts.
We should remember that this is not some offhand, experimental use of technology; here we are talking about high-stakes tests whose results are used by governments such as Australia, the UK, and Canada, among others, as well as many universities across the world, to make immigration and admission decisions.
Not so long ago, such a task would be exclusively performed by trained personnel, preferably with a background in the relevant discipline.
And, there is also the alarming realization that if bots can perform tasks such as judging the intellectual merits and grammatical accuracy of an essay, then it is only a matter of time before your skills, too, can be superseded by bots, whether you are a junior office worker or a business manager.
In short, the time has come for humans to face up to the fact that the rise of AI and the subsequent “singularity” can potentially put an end to the concept of work as we know it forever.
Thinkers such as Ray Kurzweil, Jürgen Schmidhuber, or Louis Rosenberg, have predicted that the ultimate singularity will be with us sometime between 2030 and 2050.
Singularity is a hypothetical stage in the development of computers that a unified network of AI capable of self-improvement enters a “runaway reaction,” whereby it goes through countless cycles of self-upgrading, becoming more intelligent after each cycle.
If repeated enough times, this will lead to an “intelligence explosion,” and give birth to the world’s first artificial superintelligence.
According to Kai Fu Lee, an AI expert and a venture capitalist, some 40% of all jobs around the world will be lost to AI in the next 15-25 years.
Meanwhile, well-known figures from the world of science and technology, including Stephen Hawking and Elon Musk, have gone so far as to warn the world about the potential threat posed by artificial intelligence.
If such warnings are to be believed, then is building AI opening Pandora’s Box?
This uncertain prospect raises a few questions: if a benign rise of AI is going to liberate all humans from the burden of work, then what exactly will everyone do? How will humans make a difference in the age of AI? Will people, who seem to be so happy to spend a day or two in complete idleness, stay happy when they have nothing to do ever again?
Many contemporary psychological models consider occupation as part of humans’ psychological needs. A 2011 study by Quinn M Pearson on 155 working women came to the conclusion that both “job satisfaction and leisure satisfaction were positively correlated with psychological health.” This may mean that spending a lifetime in front of Netflix without anything to do whatsoever may not be as desirable as it first appears.
But some suggest that our hunter-gatherer ancestors did not really have any jobs for hundreds of thousands of years—presumably with the obvious exceptions of hunting and gathering which took at least two hours of their lives each day, according to some estimations.
However, it is true that regarding one’s profession as part of one’s identity and self-image is a relatively new phenomenon, which has been with us since the agricultural revolution circa 10,000 BC. And, there is little evidence to show depression was particularly prevalent among our jobless hunter-gatherer forefathers.
The best answer to the conundrum at hand, however, may be offered to us by AI itself. Come to think of it, if an artificial superintelligence is just around the corner, then perhaps it can also give us “the ultimate answer to life, the universe, and everything,” to borrow Douglas Adams’s words in The Hitchhiker’s Guide to The Galaxy.
But we do not know whether a global network of unified superintelligence, millions of times smarter than all humans, will be benign because we—by definition—cannot follow its line of thinking.