Watson, a computer system developed by IBM, made headlines around the world earlier this year when it beat established Jeopardy champions Ken Jennings and Brad Rutter in two televised exhibition games.
Jennings later claimed the format of the game – which relies on quick reflexes as much as the ability to answer trivia questions – favoured the computer, suggesting that without the buzzer he would have won “every time”. However, he was impressed enough to semi-jokingly conclude: “I, for one welcome our new computer overlords.”
Until now it has been assumed that job growth will be in the information economy. This assumption has underpinned attempts to expand access to higher education (though not overall spending). However, an increasing number of technology analysts believe recent developments in computing may mean that some white-collar jobs are more vulnerable to technological change than those of manual workers. Even highly skilled professions, such as law, may not be immune.
Many established music and publishing retailers have struggled to adapt to the challenges posed by the rise of e-commerce. The American bookstore chain Borders recently filed for bankruptcy, citing competition from online retailers and e-books, prompting congressman Jesse Jackson Jr to claim the iPad – which he himself uses – “is now probably responsible for eliminating thousands of jobs”. He also demanded to know, “what becomes of publishing companies and publishing company jobs?”.
Journalists have been feeling the heat for some time. Documentary maker Tom Streithorst, who has reported on conflicts from Colombia to Libya, believes the shift from print to the web means “what used to get you a decent paying job now just gets you a minor celebrity in the blogosphere”.
Meanwhile, Larry Birnbaum and Kris Hammond of Northwestern University in Illinois, have developed software that can write short, factual articles. The resultant spin-off company, Narrative Science, already has contracts with a variety of news organisations and is moving into writing copy for online advertisements.
Earlier this year the New York Times reported on a complex lawsuit in which software was used to process 1.5m legal documents at a fraction of the time and expense it would have taken lawyers and their assistants. The software, developed by California-based Blackstone Discovery, is reportedly advanced enough to filter documents by concept as well as simply by relevant terms.
The Nobel-prize winning economist Paul Krugman believes these are not isolated examples. In a recent NYT column he states that since the early 1990s technological change has meant that “high-wage and low-wage employment have grown rapidly but medium-wage jobs – the kinds of jobs we count on to support a strong middle class – have lagged behind”. In short, Krugman predicts that “robot janitors are a long way off; computerised legal research and computer-aided medical diagnosis are already here”.
This pessimism about the fate of the middle manager and office worker is not limited to the US. Stephen Overell, associate director of policy at The Work Foundation, believes a similar trend is happening here. “There is growth at the top and bottom of the British labour market, rather than the middle, more professionals and senior managers but also more low-paying personal service jobs,” he says.
Alexander Grous, a technology specialist from the Centre of Economic Performance at the LSE, agrees. He suggests “the digital industrial revolution espouses many of the tenets that its mechanical predecessors did: disruptive; productive; facilitative; fearsome; irreversible”.
James Callander, managing director of recruitment consultancy FreshMinds, says automation in the area of data extraction is on the rise. “Many jobs in banking and insurance have been replaced by computers or bypassed as consumers increasingly use the internet,” he says.
In those jobs that remain, the impact of technology is rapidly changing the kinds of skillsets required. Callander says: “We specialise in management consultancy and there’s an increasing demand for those skills in marketing departments of large corporates, which used to be all about public relations and advertising. Now it’s all about using quantitative data to measure the impact of digital campaigns”.
Technology companies say the process of white-collar automation is gathering pace. “Although the UK is behind America [in service sector automation] it is catching up fast,” says John Everhard, technical director of Pegasystems, a company specialising in business process automation.
He also thinks the range of tasks that can be performed by software is getting broader and “starting to shift from back office automation to more complex processes, such as customer relations management, as the technology gets more sophisticated”.
Although Everhard accepts that some in the industry have in the past been too quick to make claims about the extent to which certain tasks can be satisfactorily automated, he is convinced that this is no longer the case, since “the software is finally catching up with the hype”. He points to the experiences of some of Pegasystems’s most satisfied customers, such as Kenneth Klepper, chief operating officer of US pharmacy benefits manager Medco, who claims his firm’s use of Pegasystem’s software and expertise boosted productivity by 30%.
Even when jobs are not being fully automated, software means firms can significantly expand their business without needing to hire more middle managers.
Jason Follin, project office manager for business software provider SAGE, claims that his company can help businesses organise their workflows to be organised down to the hour. He claims senior managers can now instantly see how projects are progressing, and understand where demand is coming from. “Gut instinct and human judgment are increasingly being replaced with real-time data,” he says.
However, both experts and practitioners caution that automation has the potential to backfire on firms who attempt to do things too quickly. Everhard emphasises that when automating “it is important to plan things out correctly. For instance, you should focus on the easiest tasks and work outwards, rather than trying to automate more complex processes immediately”.
Communication with staff is also important when implementing technology, especially when it involves changes to previous working practices. Follin says that “when we first moved to electronic timesheets staff were initially worried that we were going to start counting the number of tea breaks”. However, after explaining the benefits of senior management being aware of what people were doing, “people quickly embraced the system”. He warns that “if you impose change in a dictatorial fashion, people will do a sub-optimal job”.
It is also important to realise that this “hollowing out” process will not be instantaneous. Overell of the Work Foundation states that “although there is definitely strong evidence that the labour market is going through a large degree of polarisation, we must be cautious about claims that millions of jobs will suddenly disappear”. He also warns that “trends are only trends if they take place over decades, not over a year or two”. Callander similarly believes that “just as you will always have technophobes who shun innovation, there will always be those who are too optimistic about what technology can achieve”.
Even though Grous of the LSE is confident that change is occurring and that it is “agnostic of the colour of the collar”, he points out that any job losses need to be “balanced against the commensurate additional opportunities created in the process”. Trevor Ward, EMEA General Manager for AtTask, a company with more than 1,100 clients worldwide, takes a similar approach. He argues that automation will lead to the “democratisation of work”, with workers being given more flexibility and control over their task workflow as software allows them to set priorities and share work. Ward predicts that middle managers, freed from some of their more routine tasks, will have time to engage in more strategic thinking.
Interestingly, many of those in charge of developing software to automate tasks previously performed by humans have mixed feelings about the impact of their work. Birnbaum is at pains to emphasise that Narrative Science has “been working very closely with journalists in developing this technology and others”. He concedes that it is going to be a big challenge “to organise ourselves socially and economically in the face of massive intelligent automation”.
However, short of rejecting technological progress, which would have negative consequences for society, there is little that can be done to stop these changes. Birnbaum points out that, “since these technologies are capable of producing great wealth it would be a terrible waste not to utilise that potential for human benefit”. Unfortunately, this may be little consolation for those who have been made redundant, or whose jobs are threatened, by the creation of ever more intelligent software.