The recently concluded World Artificial Intelligence Conference in Shanghai featured a highly anticipated debate between Elon Musk and Jack Ma on Artificial Intelligence. Musk, has long been a vocal commentator on AI and its associated risks, and pretty much had the same talking points that he has been reiterating for a few years. And while the ‘face-off’ didn’t deliver the ‘fireworks’ that many expected, during the nearly 45-minute conversation Jack Ma made four important points about the coming ‘AI age’, some of which seem to diverge from his previous public statements on the interplay between AI, jobs and ideal working hours.
This is what he said: First, advances in medical sciences will increase life expectancy to around 120 years. (As per WHO, current global life expectancy at birth is approximately 72 years.) Second, there will be no shortage of jobs, in fact, there won’t even be a large need for jobs. (This is, notably, at odds with his statements at the 2018 World Economic Forum, where he lamented that AI will “kill a lot of jobs”.) Third, jobs assisted by AI will only require 12-hour work-weeks. (He has previously advocated the infamous spirit of 996, that is, 9 am to 9 pm, 6 days a week). The fourth, and perhaps the most important, that education needs to be re-engineered to prepare for this world order.
As incredible as a 156-hour weekend sounds, let’s first look at some available estimates over the next 30 years to get a better sense of things. Despite declining fertility rates, by 2050, the Earth’s population is expected to be around 10 billion. Much of the growth is expected to come from the 47 least developed countries. Various industry estimates peg AI’s share of the global economy at around 20 per cent or $50 trillion with an investment of around $5 trillion.
With continued advancement in today’s specialized/narrow artificial intelligence capabilities through repetition and reinforcement, expectations are already high that jobs involving high degrees of repetition, whether physical or cognitive, are ripe for displacement. For example, ‘Robo reporters’ like Bertie (Forbes), Heliograf (Washington Post) and Cyborg (Bloomberg) are already being used to cover sporting events, elections and financial reporting. Also consider that technology advancement largely enables the replacement of labour with capital. And that for most people, the only ‘private property’ that they can exchange for compensation is labour.
There are some counter arguments here though. One is that the Industrial Revolution, which was expected to displace most jobs that were created as a result of the Agricultural Revolution, actually ended up creating more jobs. And, just as we have seen a number of new jobs/careers opening up since the proliferation of the Internet, we will continue to see new types of jobs being created in the market. But James Barrat, author of Our Final Reinvention, makes the important distinction that the Industrial Revolution did not simultaneously impact multiple industries at the scale that it is likely to happen now. And that the impending replacements are creeping higher up the employment value-chain and happening faster, making short/medium term re-skilling all the more challenging.
So, when you consider that mainly ultra-high skilled and Intelligence Augmented jobs will be left in the market, a 12-hour work week/156-hour weekend certainly seems plausible. The problem with this scenario is that a 168-hour unemployment week may be more common, exacerbating the already lopsided income distribution situation. This sounds grim enough without considering that people could also live up to 50 years longer than they currently do.
Society needs to start preparing itself for this shift. The ongoing debates around Universal Basic Incomes are a starting point, but these often hit a stumbling block over what is the appropriate amount that will let people live comfortably yet motivated enough to pursue creative endeavours. Too little, and it will not solve the problem. Too much, and the popular fear narrative is that recipients will turn into stay-at-home gamers, slowly degenerating to resemble the ‘Eloi’ described in H G Wells’ Time Machine (future human descendants who have lost their curiosity and ingenuity and regressed into beings that only play, feed and reproduce).
Some tweaks have been suggested such as a form of tiered basic income where ‘contributing’ members can receive a higher compensation though it remains unclear how to define contribution, and how much scope there will be for it. There are also, rightly, concerns over how such an exercise can be sustainably funded. Most countries, even today, cannot afford to provide a basic income for a non-working majority. On top of this if you factor in that the big tech companies best positioned to reap profits have structured themselves in ways to minimise tax payments (eg. setting up their HQs in low-tax jurisdictions etc.) and do not actually pay taxes in many of their markets, it remains unclear where the funding will come from.
Nevertheless, politicians around the world, are slowly beginning to talk about the idea of a basic income. When asked by the UBIE (Unconditional Basic Income Europe) movement, a number of candidates for the European Parliament advocated in favour of it. Andrew Yang, an entrepreneur, centered his campaign for the Democratic party nomination on it. Finland actually ran a basic income trial which concluded earlier this year. Even the Indian National Congress’ proposed NYAY scheme was a nod to the concept of a basic income (though, it was not universal). And, if a significant number of jobs are impacted, an Overton window may open up.
But a basic income by itself cannot be considered a panacea. Re-engineering the way society learns is crucial to ensure that it can cope with this shift. An education system based largely on rote learning and reproducing facts, which is already waning in relevance considering the poor employability of most graduates, will do little to prepare its products for a world where only ultra-high skilled jobs remain. Futurists recommend an emphasis on STEM, humanities, adaptability and application of thought but the reality is that no one has concrete answers today. As Michael Jordan, UC Berkeley Professor and one of the leading voices on machine learning, puts it, when it comes to Artificial Intelligence“scientists are often as befuddled as the public.” Whether such a scenario presents itself in 20, 50 or even 100 years is not relevant at this point. It is inevitable and possibly a more urgent problem to solve than the fear of ‘Skynet’ taking over.
(Prateek Waghre is a student of the Technology and Policy Programme at Takshashila Institution)
The views expressed above are the author’s own. They do not necessarily reflect the views of DH.
