Many of us have already lost the 「race against the machines」 — we just don』t know it yet. That is the conclusion of new research by the Organisation for Economic Co-operation and Development.
Unlike most studies into the impact of automation, this one does not rely on informed guesswork about what machines will be able to do in 20 years』 time. Instead it takes three core skills that three-quarters of us use every day in our work — literacy, numeracy and problem-solving with computers — and compares our performance against the abilities of machines. The results are sobering, but rather than a reason to despair, they suggest we might want to rethink the race altogether.
The OECD has a good idea of our proficiency in these areas because it has put 216,000 adults in 40 countries through a 50-minute assessment called the Survey of Adult Skills. In the survey a group of computer scientists was given the same test and asked which questions computers could answer, using technology that exists but has not necessarily been rolled out yet in the workplace. The conclusion? Almost a third of workers use these cognitive skills daily in their jobs and yet their competency levels have already been matched by computers. About 44 per cent are still better than the machines. The remaining 25 per cent have jobs that do not use these skills every day.
There are two caveats. First, the OECD only asked computer scientists how well they thought machines could do. The results would be more compelling if machines were actually put to the test. Second, just because technology exists does not mean it will be deployed quickly in the workplace. It depends on how easily it can be made operational, how much it costs relative to the value it creates, and whether companies have the appetite to invest.
And yet, the implications of the study are hard to shrug off. Stuart Elliott, the author, concludes that in 10 to 20 years, only workers with very strong literacy and numeracy skills will be comfortably more proficient than computers. At the minute, only about one in 10 working-age adults in OECD countries are of this standard.
It is true that the education systems in most countries have been raising their game: younger people tend to have better skills than older people (the UK being one notable, and worrying, exception). But even if you take the most skilled generation in the most skilled country — young people in Finland — two-thirds still do not meet these top levels of literacy and numeracy. Short of astonishing improvements in education, it looks like only a minority of people can win this race.
But that does not necessarily mean everyone else becomes redundant. In most jobs, people combine cognitive skills with other human abilities: physical movement; vision; common sense; compassion; craftsmanship. On many of these fronts, computers are behind humans, if they are in the race at all.
The risks to workers from ever smarter computers are clear, but the opportunities will lie in maximising the value of their human skills. For some people, such as talented chefs, the battle is already won. Others might need to harness the computers to leverage their human talents. For lower-skilled workers, there is already evidence to suggest that working alongside technology can help their prospects.
Research by Richard Blundell, an economics professor at University College London, suggests the low-skilled tend to fare better in big companies that invest heavily in research and development. They have higher wages than other low-skilled workers and tend to stay with their employers for longer.
These findings, though preliminary, are a reminder that technology does not necessarily mean doom to all but the highest skilled. The best response to the race against the machines is not to hold the machines back; it is to help the humans jump aboard.