In 2013, Oxford economist Carl Frey and Machine Learning expert Michael Osborne found that there is a high chance that engineers will find ways to automate 47% of jobs in the United States, possibly within “a decade or two”.³¹ What is significant about their findings is that unlike previous technological revolutions this one will impact workers across the spectrum, from low-skilled to white-collar jobs. The potential scale of the disruption created by technological developments, such as artificial intelligence, machine learning and big data, requires that governments think deeply: how can they help mitigate the risks and fully harness the opportunities?

Technological innovation in recent years has made computers, robots and software so sophisticated that machines are now entering the realm once thought to belong exclusively to humans: cognition. Computers today can recognize patterns and generate insights being used for fraud detection, medical diagnostics, legal research, and auditing, among others. Artificial intelligence algorithms can process thousands more documents – and then act faster – than any human and are free from human biases. Their productivity also need not be interrupted by rest breaks or lapses of concentration. The new technological revolution will create tremendous societal benefits – the creation of new goods, services, markets and jobs, greater productivity, etc. 

The Centre for Strategic Futures, a think-tank within the Singapore Government, and the Ministry of Manpower in Singapore, are conducting a study similar to Frey and Osborne’s using Singapore labour data, to explore some of these issues. One interesting question the study has raised so far is whether some professions might face “broken career ladders”, where entry-level workers no longer have a clear path for career progression because the tasks they would traditionally perform in order to progress have been automated. For instance, entry-level tasks in professions such as law and accountancy, e.g. basic research and cleaning up financial data for processing, can be automated more easily than higher-order cognitive tasks, such as framing and solving problems or making judgements, performed by more senior professionals.  

The traditional approach of helping workers upgrade by “up-skilling” will not necessarily reduce a worker’s susceptibility to being displaced by the new wave of automation. Workers will need to develop new skills to take on very different kinds of jobs, possibly in different industries. Presently, most governments dedicate resources towards helping low-skilled workers secure better jobs through training and education. Yet this shift will affect workers across the employment spectrum. Thus, governments need to work with stakeholders to rethink the kind of pre-employment and post-employment training institutions should offer to enable professionals to keep pace with these developments. What are the new job opportunities that may emerge? How can we ensure that the benefits accrue to a broad spectrum of society and not only to the most highly skilled and well-resourced?

Governments need to consider the role they can play in promoting – or stymieing – the use of technology and automation by industry. 

For example, the automation of cognitive tasks can transform auditing, allowing for real-time audits. Current auditing rules in some countries require all suspicious transactions spotted during an audit to be investigated in detail. Such a rule works today because human auditors pick only a sample of documents to audit, rather than going through their clients’ entire database of documents, which results in a manageable volume of suspicious transactions to investigate. However, because it is likely to throw up many more suspicious transactions and impose an overly onerous burden on both the auditor and its clients, having such a rule means auditing firms are unlikely to use algorithms for their audits even though they can potentially make a chief financial officer’s job easier by providing real-time analysis and insights into an organisation’s financial health. What regulations impede and encourage the adoption of automation, artificial intelligence and/or analytics? On a global scale, will there need to be a set of international standards to encourage and manage the impact of automation, given the risk of arbitrage? 

How a government approaches the ethical and legal implications of technologies like autonomous vehicles (AVs) would also influence how widespread the adoption of technology and automation will be and the pace of its adoption. AVs present the opportunity to radically redesign mobility solutions and also create new jobs in a new industry, but the autonomy also raises questions about which party should be liable in an accident – the manufacturer, software developer, the owner or the passenger in the AV. Should governments be bold in encouraging innovation, whilst helping the “losers” take part in the broader improvements – or at the least helping to buffer them from the downsides? 

At times, it may seem as if technology is a force greater than humans, forcing workers and businesses to adapt – or perish. Yet governments play a key role in shaping how technology advances. The sooner governments, in partnership with the rest of society, examine the future impact of this structural shift, the sooner they can act to ensure the shift benefits society.