The easiest thing for policymakers to do is to bury their heads in the sand - preferring to maintain the status quo rather than trust what they see as the ‘Mystic Meg’ guesswork of researchers trying to predict the future. They fear getting it wrong. Backing the wrong horse. Or in this case, the wrong driverless car.

But denying the future will disadvantage those who are most vulnerable - those who don’t have family to fall back on if they need to retrain, or who are in jobs with few employee protections.

In a previous piece of research, in partnership with Pearson, Nesta looked at the impact of trends including automation, the growing ‘green economy’ and changing demographics on the workforce up to 2030. We predicted that around 10 per cent of workers are in occupations which are expected to grow their share of the workforce, whilst 20 per cent are in occupations that are expected to shrink.

The Government’s answer to these shifts has been to ‘invest in digital skills’ but this isn’t enough. Not all digital skills are created equal, and we wanted to understand more about which ones would truly ‘future-proof’ your skillset, and which skills were in fact more likely to be needed in jobs that could disappear.

So, last week, Nesta published our newest piece of analysis on future employment and digital skills, which looked at 41 million jobs adverts to identify the digital skills currently needed in those jobs most likely to grow by 2030 and those in the jobs most likely to disappear. What we found was perhaps even more stark a divide than we - at least I - had expected.

That’s because there are jobs with buoyant prospects that don’t need many digital skills at the moment - including teachers and chefs. Where digital skills are needed, they are noticeably different in jobs likely to grow and jobs likely to decline. What sets ‘future-proof’ digital skills apart is their use for non-routine tasks, problem solving and creation of digital content.

In short, if you are just inputting data it may not be long before a robot can take your place. But if you are creating something with that data, your job is not only less likely to disappear but we predict that it will become more important. Skills in animation, multimedia production and design engineering are also seen in jobs with buoyant job prospects, whilst clerical skills like typing and invoice processing performed less well.

This research isn’t just relevant to policy wonks like me. Government, business people, educators and individual workers also need to sit up and take note.

In schools, kids should continue to be taught both advanced digital skills like coding, as well as the basics of using digital software to communicate, design and create.

Our original research with Pearson also highlighted the increasing need for soft and cognitive skills. Teaching these might be tough - and it’s even harder to test for them - but that will have to be a priority too.

Government has recognised that we need to overhaul current careers advice in schools, but we also need to be able to map how jobs are changing to keep this advice current. Open Jobs is a Nesta initiative that will aim to offer real time labour market data to support all kinds of decisions, including helping careers advisors and even individual students.

The lessons are much the same for employers and individual workers. We aren’t suggesting that companies retrain all secretaries as animators, but instead advocating for a learning culture where people are able to understand what new skills might be able to supplement those they already have, and how to acquire them.

For policymakers, it is a reminder to look beyond the ‘digital’ or ‘tech’ label and consider the specific skills that the workforce of the future will need. And, of course, to actually fund training. With the number of self-employed people on the rise (and those people about half as likely to participate in work-related training), alongside an ever-growing need for lifelong learning, it seems like an important moment to invest in retraining as well as school and tertiary education programmes.

One consistent message coming out of our work is that policymakers, teachers, and the workforce need to stop fearing the future. Or at least they should recognise that they have an opportunity to shape it. And, most importantly, we musn’t shy away from skills that seem difficult to teach - it is these that set us apart from the robots.