The jobs at the end of the universe

Researchers say technology will make all but six human skills redundant. Douglas Board asks whether they are right

Good news – six human skills will survive the onslaught from ever smarter machines. But the bad news is that the intelligent computers will surpass us in all the rest.

MIT researchers Erik Brynjolfsson and Andrew McAfee have looked into the future* and report that the achievements of smart machines are about to explode.

Apple’s new personal assistant, Siri, vastly outpaces what seemed possible only 10 years ago. In 2010, Google announced that its fully automated vehicles had driven more than 1,000 miles on US roads without human intervention – something declared “non-automatable” in a carefully researched study by economists in 2004.

And last year, Watson (an IBM supercomputer) beat two human champions at the American TV quiz game Jeopardy!. The quiz gives contestants answers and requires them to come up with questions. To succeed means going beyond the recognition of fast-changing physical patterns required to navigate a vehicle but also mastering human quirks such as trivia, word-play and puns.

Mr Brynjolfsson and Mr McAfee, director and associate director of the MIT Center for Digital Business respectively, point out that these are far from isolated examples nor a matter of the exceeding of ignorant expectations.

Moore’s Law – the proposition that computer capabilities double approximately every 18 months – suggests that we are riding an exponential tidal wave that will, say the MIT researchers, make many human skills redundant.

But six will survive, say Messrs Brynjolfsson and McAfee, no matter how fast and smart computers become. Those skills are: statistical insight; managing group dynamics; good writing; framing and solving open-ended problems; persuasion; and human nurturing. These will define the jobs they think will exist at the end of the universe.

So is it time to focus on these six skills, to make ourselves future-proof?

Lacking a crystal ball or time machine, we can only try a lateral thought experiment instead: how do the six measure up in the boardroom? At the top of organisations ruthless delegation is usually standard operating procedure: if it can be done by cheaper people – or smart machines – it will probably be delegated, rather than prized.

Taking two of the six skills, I can say that in my experience the products of good writing and human nurturing (if we mean by that the nurturing of the boss!) are valued in the boardroom, but within limits.

I remember a newly appointed chief executive salivating over the quality of the speech he was going to deliver on a significant industry occasion. Winning the chief executive role justified him hiring a top wordsmith, for which he was happy to pay the going rate. But he wasn’t about to put the writer in charge of a piece of the business or give him stock options.

Similarly, top secretaries know that while mopping the boss’s brow is valued, the value won’t be a percentage of the company’s equity. Good writing and human nurture may well be jobs left standing at the end of the universe, but they probably will not be the best paid.

Our next three to be tested – framing and solving open-ended problems, managing group dynamics, and persuading – are clearly qualities prized in senior executives, usually called strategic thinking and leadership, respectively. So those make sense as bets for the end of the universe.

The remaining skill – statistical insight – stands out as different, and therefore interesting. Messrs Brynjolfsson and McAfee dispute that because computers can crunch ever-larger warehouses of data faster and faster, fewer people will have to become slick at statistical analysis. Instead, they think smart statistical interpretation will become more necessary and more sexy.

They might be right, but so far as boardrooms are concerned, we are not there yet. Looking at my clients as well as at successful candidates from 18 years in executive search, only a few managers had an awesome grasp of numbers and their implications. Their ability was as likely to be photographic recall of the numbers in a spreadsheet as analytical prowess.

I have found three more common quantitative abilities to be valued at senior levels: making the meaning of numbers come alive either visually or in words; a keen sense for when numbers should be an important part of a story yet are missing; and not being bullied by impressive correlations into assuming causality.

Indeed, when we view the two researchers’ six skills from the perspective of the boardroom, what appears strikingly absent is any reference to taking decisions.

While senior managers do take decisions which “solve problems”, usually when the stakes are very high (fire Fred, exit Asia), they take decisions at least as often which re-shuffle or re-define problems, or even create them.

This is because many decisions are not solutions but a choice between different headaches.

The decision by Paul Polman, chief executive of Unilever, to commit the group to doubling in size within its existing environmental footprint, is an example of this. Instead of solving anything, it creates headaches which did not exist before. The gamble is that some of those headaches – for example, can we create and popularise a water-free shampoo? – will prove productive.

So perhaps humans will have a broader function than that described by Messrs Brynjolfsson and McAfee. Their emphasis is on the word “skills”.

Skills live comfortably alongside defined goals that can be achieved, or problems that can be solved.

But they do not spring to mind when we want to describe an action which seems closer to a choice, or an assertion of values – such as whether to have soup or salad as a first course. Much top management decision-making is more like the latter, but less pleasant.

Before solutions, before problems, before skills, come wants and values. At different times and in different places we find actions or experiences enjoyable, interesting, repulsive or boring (an expression of wants), and we also evaluate those experiences as meritorious or despicable, substantial or trivial – an expression of values.

At the end of the universe we will still be playing games with each other, some of them violent, many of which only smart machines will make possible.

Which problems we can solve better than smart machines will matter far less than which group of entities cares enough to debate what “better” means.

Watson winning Jeopardy! is, as Brynjolfsson and McAfee explain, a tiny foretaste of things to come. But can we say Watson wanted to win?

What is distinctively human is not winning Jeopardy!, or even having invented it – but to have invented it and to have invented wanting to win it – and finding the whole thing utterly trivial, all in one go. For good or ill, skills aren’t really what we do.

*Race Against The Machine (Digital Frontier Press 2011).

Dr Douglas Board is a senior visiting fellow at Cass Business School. He advises on career change (

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't cut articles from and redistribute by email or post to the web.