During last month’s historic game of Go at the Four Seasons hotel in Seoul, international champion Lee Sedol burst into an exuberant grin after his first win against Google’s artificial intelligence (AI) program, AlphaGo. It would be his only win against the program in the best-of-five match.
While Chess can be won through simple but extensive calculations, Go is said to be exponentially more difficult due to its many more possible moves, and people like Elon Musk and AI researcher Remi Coulom noted that a win for an intelligence program was predicted to still be years away. The victory of AlphaGo raises questions on the state of artificial intelligence and its future role in society.
As AI develops, students and employees increasingly begin to fear for the potential of AI to endanger their chosen careers. We only need to look at history to observe how industries have been displaced by automation and the mechanisation of labour. The printing press supplanted the painstaking work of scribes in monasteries. However, many people would note this temporary unemployment was ameliorated by the generation of other industries – from booksellers to librarians to an influx of writers.
Contention still exists around whether technological advancements will lead to long-term unemployment. Professor Mark MacCarthy of Georgetown University suggests that machines will increase opportunities for employment, while studies at the University of Oxford suggest the opposite. The latter trend seems evident – where workers were once predominantly farmers, humans have now been rendered obsolete in an automated agricultural industry. More obviously, automation is present in our everyday interactions, from buying that shameful packet of instant noodles from the self-checkout machine, to the annoyingly pre-programmed responses of call centres.
In the same way that my grandma couldn’t have comprehended Pewdiepie’s career, technology may create jobs we can’t even imagine right now. But it could also create programs so intelligent that we seem as useless as a stone tablet next to an iPad.
But surely computers are only good at repetitive jobs, letting us pursue something interesting, or creative, something uniquely human? Our degrees in Renaissance art or linguistics or architecture are safe. And yet expert musicologists can no longer distinguish pieces composed by programs from those composed by human musicians. Similarly, white collar jobs in stock trading and medical diagnosis are already partially mechanised. Not only does automation have the potential to overtake manual labour, but, given enough time, all other industries as well.
A potential solution to the automation of jobs seems to be an increased focus on science, technology, engineering and maths (the so called STEM) subjects and courses, redirecting career channels into areas that actually revolve around the development of AI and technology. If you can’t beat them, join them. The University of Sydney seems to acknowledge this, noting on its STEM webpage that 75 per cent of the fastest growing occupations require STEM skills and knowledge, planning to enforce mathematics as a prerequisite for 62 of its degrees from 2019. However, the deficit in STEM students continues, as public sentiment still vilifies a career in science as cold, or removed from society. Australia’s Chief Scientist, Alan Finkel, notes the decline in interest in STEM subjects in high school, leading to a shortage in tertiary enrollments and participants in the workforce. A career in IT is clearly not the new ‘IT’ career.
The University seems to echo the government’s model of hailing STEM as the future. The introduction of the ‘Ideas Boom’ rhetoric and pursuit of an innovation-based agenda clearly seeks to push STEM to the surface, with the government committing 12 million dollars to maths resources and programming subjects in schools.
Despite this renewed focus, the government has neither internalised nor acted on its supposed support of STEM subjects beyond these shallow measures. It continues to divest funding from climate research and the CSIRO, instead channelling it into dubious projects like pseudo-scientific research into health impacts of wind turbines. And while the aggressive innovation agenda certainly may give STEM careers the status they were missing, this may further isolate an essential ingredient in encouraging young people to pursue STEM: a sense of curiosity revolving around the romance of scientific study. Indeed, the CSIRO has been forced to cease research with no direct link to jobs and economic growth, severely limiting the potential innovations that so often tangentially emerge from such research. Other obstacles are present to promoting innovation, including the extremely disturbing underrepresentation of women in STEM fields, where they occupy only 16 per cent of workers, and number only a third of the number of men in the highest income bracket.
Perhaps the government is actually more strategic than they are given credit for and is preventing further technological innovation and the potential long-term unemployment that may come with it. It’s difficult to determine the more hazardous risk: automation and loss of many viable careers, or the climate change induced global catastrophe that would occur under the Australian government’s ironically short-term outlook.
Clearly the government, and society itself, needs to be even more stringently future focused. Without a reflection of this engagement in science in all aspects of policy, we could find ourselves overwhelmed by AI. Automation of labour has the potential to collapse the capitalist economic model that dominates our societies. Without work –that is, with unlimited leisure time – our social, economic, religious, political and scientific paradigms will be completely overturned. There is no denying the drastic effect artificial intelligence will have on our lives.
And yet we still underestimate the potential of AI. Stephen Hawking noted that the technological singularity, a point where AI would be capable of autonomous self-improvement, “could spell the end of the human race”. AI that improves itself would therefore get better at improving itself, leading to the singularity level of intelligence as soon as 2040, according to philosopher Nick Bostrom.
Concerns and postulations regarding AI have also captured audiences beyond academic communities, from Stanley Kubrick’s 1968 film 2001: A Space Odyssey to a far more romantic rendering in Spike Jonze’s 2013 film Her. These films explicate very different expectations of AI, but share a similar message: whether as commander of a spaceship, or a personal assistant operating system, technological advances in artificial intelligence will completely alter the roles of human beings, and industries available to the workforce.
It’s difficult to tell what will happen in the future. While technology and therefore our ability to predict the future has improved, the rate of advancement has also increased exponentially, as evidenced by AlphaGo’s unexpected victory over Sedol. We can’t know how the development of artificial intelligence will affect job markets and what roles we could effectively perform in the future. We don’t know whether our careers are protected or will be protected by governmental policy and whether its innovation will lead to more jobs or fewer. Que Sera Sera?
For now, we have internalised the commencement speeches of Steve Jobs and J.K. Rowling, and platitude-filled career guides, believing that our purpose in life is to pursue our dreams and passions. To never give up, because #YOLO. We try to avoid becoming unsatisfied by our future careers.
Ironically, as C.G.P. Grey noted, we try to avoid becoming mechanistic, soulless robots.
Robot art by Johanna Roberts.