Skip to content

Creativity versus robots

Jane Jacobs famously pointed at creativity and diversity as engines of growth and prosperity, noting that innovation flourishes in diverse economies that facilitate interactions between creative people.

Such interactions may be relevant to explain key events in economic history. These include for example the industrial revolution, where the idea behind Richard Arkwright’s water frame came from John Kay, who was working on a similar machine with his neighbour – the inventor Thomas Highs. Or the digital revolution, where the iPhone combined Apple technologists’ capabilities in hardware and software with Sir Jonathan Ives’s design ingenuity.

Today, more people work in creative jobs than ever before. In the UK, there are 1.8 million creative jobs according to government classifications based on our Dynamic Mapping method, growing roughly three times faster than the workforce as a whole.

Broader conceptions of creativity give even higher numbers, with Richard Florida famously estimating that one-third of the US workforce is now in creative occupations, accounting for about half of all wages and salaries. In 1980, at the dawn of the computer revolution, this ‘creative class’ made up less than 20 percent of the workforce, up from around 10 percent at the turn of the 20th century.

Many jobs we consider creative today may not be so tomorrow if they fall within the scope of automation

To be sure, people in creative occupations have vastly benefited from the computer revolution, as digital technologies allow people that are in the business of selling creative ideas to transmit these to about anywhere in the world almost instantaneously. Nevertheless, many jobs we consider creative today may not be so tomorrow if they fall within the scope of automation.

In a recent study, two of this post’s authors assessed the prospects of automation for the US workforce over a 10-20 year horizon. They concluded that 47 per cent of jobs in the current workforce were at high risk of computerisation. This has led others to speculate on the equivalent ‘vulnerability’ of workforces in European countries. This research raises several questions, including how resistant are today’s creative jobs in particular to automation and, given international variations in creative employment, what does this imply for different countries?     

It also raises the question of what exactly is creativity? According to Oxford Dictionaries creativity is “the use of imagination or original ideas to create something.” At the heart of our definition is therefore an element of originality. Crucially, for our purposes, O∗NET – an online service developed for the US Department of Labor – provides open-ended descriptions of specific tasks associated with various occupations, as well as a host of variables objectively ranking occupations according to the mix of knowledge, skills, and abilities they require.

Using this data, we hand-labelled 120 US occupations as creative or non-creative, asking the question: does this job require the use of imagination or original ideas to create something? For classification, we have been developing an algorithm to estimate the individual probabilities of all 702 occupations in the US being creative given a previously unseen vector of variables derived from O*NET.

Our current results suggest that 21 per cent of US employment is highly creative – that is, has a probability of more than 70 percent of being creative (a much smaller fraction of jobs in the United States than estimated by Florida). These occupations include artists, architects, web designers, IT specialists and public relations professionals. Furthermore, we find only a very small share of the creative workforce to be susceptible to automation: 86 percent of workers in the highly creative category are at low or no risk of automation.

In a second step, we replicate the analysis for the United Kingdom. Relative to the United States, the UK has a higher fraction of creative employment, constituting around 24 per cent of the workforce, and of these jobs, we find that 87 percent are at low or no risk of automation. As expected, given the wider scope of creativity used in this analysis, the estimates of creative employment are bigger than the official estimates.[1]

Places that have specialised in creative work are most likely to prosper in the 21st century

While some creative occupations are susceptible to automation, people with creative skills have clearly been the beneficiaries of the digital age, as digital technologies are making creative skills even more productive. As a result, places that have specialised in creative work are most likely to prosper in the 21st century. In this regard the United Kingdom is in a good starting position – even better than the US it seems. We’ll be debating these issues and exploring the implications for policy at our Creativity vs Robots session at Future Shock on 14 November.


[1] We also checked the computerisability of each of the 30 occupations classified as creative in the government estimates – on average they have a probability of only 0.29 of being automated.

Photo Credit: gruntzooki via Compfight cc

 

Author

Hasan Bakhshi

Hasan Bakhshi

Hasan Bakhshi

Executive Director, Creative Economy and Data Analytics

Hasan leads Nesta's creative and digital economy policy and research.

View profile

Carl Benedikt Frey

Mike Osborne