“The goal is to become HBO faster than HBO becomes us.”
So proclaimed Ted Sarandos, chief content officer of Netflix, in 2013. How did Netflix do? Each binge-watcher will have their own opinion, but it’s probably safe to say that Netflix and HBO have become a bit like each other in the intervening four years: Netflix produces top-quality prestige TV, while with its Go and Now services, HBO has adjusted somewhat to Netflix’s watch-it-whenever model.
Much has been made of late about the prospect of automation replacing jobs — including both low-skilled work like transportation and high-skilled professions like the law and some forms of medicine. Ambitious remedies have been proposed to the devastating loss of income, benefits and social status that will result — including, most notably, the idea of a Universal Basic Income.
This response is intuitive: if employment is threatened, societies should take steps to ensure that the effects of unemployment are mitigated. (And it will be interesting to see whether right-wing media stop targeting “benefits scroungers” and “welfare queens” when the majority of their readers will be on benefits or welfare.) But there is a less explored, yet perhaps more likely prospect: much like Netflix and HBO, human and robot workers may simultaneously become more like each other.
One side of this equation – robots gaining human levels of intelligence and skill – has been getting all the attention, perhaps because nothing makes for a catchier headline or a more watchable YouTube video than the remarkable development of machines acting like humans (except for animals acting like humans, of course). Yet if we are increasingly accustomed to seeing robots as “human-like”, we have not yet begun to consider what it will mean to see humans as “robot-like”.
Put another way, whereas much discussion has centred on what will happen when human jobs are replaced, I am equally – and more immediately – concerned with what will happen when human jobs are seen as replaceable. If and when a majority of humans find themselves out of work, it is reasonable to expect shifts in policy to accommodate them. But if majority unemployment is the destination – and perhaps even a destination we can be reasonably optimistic about, if majoritarian policymaking keeps pace – the journey to that point is likely to be considerably more painful.
It is not just the well-documented fact of non-existent wage growth that will come to characterise this doomed battle to stay employed and employable in the twenty first century. Besides a stagnant salary, in the context of looming automation having and keeping a job will mean losing a raft of other benefits traditionally considered sacrosanct to employment.
The most obvious example is the so-called gig economy. Not only Uber but also its “nice guy” rival Lyft recruit their drivers as independent contractors, avoiding the obligation to provide any form of pension, paid annual leave or other statutory rights guaranteed to employees – while also imposing new financial burdens such as the requirement to buy new vehicles at regular intervals.
The Uber/Lyft model “works” due to the ubiquity of smartphones, which enable location tracking, as well as sophisticated algorithms for connecting drivers with riders. True automation, with the rise of self-driving cars, will likely have a far more transformative impact, removing the need for drivers altogether. But for all the major strides made in this field, even the most optimistic observers don’t expect fully automated transportation network for a decade at the very least. For today’s 35-year-old Uber driver, that means at least ten years of exploitative labour at the apex of his or her earning potential, with little ability to prepare for technological redundancy, either by saving or building a new skill set.
Or take the example of health insurance in the context of traditional employment. For all the understandable backlash against congressional attempts to slash Medicaid and throw people off insurance, a little-known bill, HR 1313, quietly passed out of committee in March. This bill would allow employers to fine employees for refusing to take, and share the results of, genetic tests which can predict the onset of hereditary diseases like Alzheimer’s and Huntingdon’s. Some protections would remain against the discriminatory use of this data by employers, but the onus would be on employees to demonstrate probable cause that that promotion they didn’t get was the result of their long-term mental health prognosis – likely a tall order in court.
Besides the fact that recent innovations have drastically cut the price of genetic testing, this example doesn’t seem to have much to do with technological change per se. Yet it is emblematic of a broader attitudinal shift towards seeing workers as eminently replaceable – whether by fellow humans (including Chinese workers and Mexican immigrants, as Donald Trump would hasten to add – racial resentment being an easy sell) or, far more significantly, by increasingly capable machines.
Even before it replaces us, then, automation reshapes us. We get paid less money, offered fewer benefits, and are asked for more and more, up to and including our most sensitive information – literally, the data that defines us as human. In the race to make them more and more like us, we humans risk becoming more and more like machines: defined by a function, judged by our ability to perform that function, and unceremoniously dispensed with when we can no longer do so. It’s a race with very few winners — and most of them live in a sunny corridor of the California coast.