David Willetts: Worrying about the effect of AI on future jobs may be distracting us from real challenges

David Willetts is President of the Resolution Foundation and a member of the House of Lords.

 There is increasing concern that AI could take our jobs.

And yes, new technologies do change the jobs we do. When I started my career as a civil servant in the Treasury in the early 1980s every division had its own typing pool. There were still almost 250,000 coal miners. People were worrying where the jobs would come from, especially when unemployment hit 12 per cent with more than 3m unemployed. But new jobs were created – though l don’t recall forecasts that they would be software engineers, social media influencers, personal trainers, or drone operators.

We are told that this time it will be different. AI is a pervasive general-purpose technology and maybe its effects on jobs will be overwhelming.  Already we can see certain cognitive jobs going – legal assistants ploughing through documents, reporting on the weather or the stock market, operating helplines.

The jobs that are affected do indeed look to be different this time.

This is down to the Moravec paradox that it is easier for a computer/robot to do intellectually hard things, such as playing chess as well as a grand-master, than to do what we assume are easier things – picking a packet from the shelf and pouring cereal into a bowl. The jobs that are most threatened are more cognitive jobs leading to another round of critiques of the numbers of graduates.

The graduate premium does appear to have fallen a bit, mainly because the minimum wage has pushed up the incomes of non-graduates who tend to be down at the lower end of the pay scale. But a graduate in their twenties is likely to earn about £31,500 compared with £26,500 for a non-graduate. That is a boost to earnings for graduates of about £450 a month. But then we have to allow for the repayment of student loans at 9 per cent of earnings above £25,000, a deduction of about £50 per month for the average young graduate out of an income of over £2,500 per month.

Most of the salary gains stay with the graduate and about 10 per cent of the gain is taken as repayment for the cost of the education that got them there. Not a bad deal.

But AI has reignited fears about the jobs that graduates do.

Some graduate recruitment may be cut back as the easiest way to adjust to fewer cognitive jobs is to stop recruiting for them. But overall the evidence is still that graduates are far more likely to be working than non-graduates. Going to university is a very good protection against economic inactivity. The percentage of graduates on out-of-work benefits does not exceed 3 per cent in the fifteen years after their graduation. For non-graduates the peak is 17 per cent before falling to 11 per cent. Given the concern about inactivity it seems odd that the political class has taken against higher education when it is one of the best tools to deal with it.

We are told these graduates are “over-educated” but there are not two neat categories of jobs. Teaching did not use to require a degree – were the graduates who went in to teaching then over-educated? And the real threat from AI is not that it takes your job, it is that your job is taken by someone who understands how to use AI to do your job better.

The danger is that this fear of AI takes us away from real challenges.

If anything we need more economic change and dynamism. The rate of growth and contraction of different business sectors is unusually low not high. And the same for job moves which are also unusually low. One of the reasons why Margaret Thatcher was wary of apprenticeships was that they trapped people in specific types of jobs in established industries with lower job mobility. That sense of the stable long-term job in a particular industry is one reason why apprenticeships are so popular but the promise of stability is hard to deliver in a modern flexible economy.

There is a growing problem of economic inactivity as Kemi Badenoch argued in her speech last week.  It is concentrated amongst the less well educated. The worst blow is just to miss good GCSEs. Doing something about the destructive cycle of resits afterwards is one of the best things we could do to get more teenagers into worthwhile training and jobs.

There have been many programmes trying to get young people into work so we can compare them and see what works. The most effective offer constructive advice and mentoring to the individual, finding out what their aptitudes are, pointing them to jobs and training programmes that best match them. The good news is that young people often have common-sense career aspirations – to be a chef or a teacher or a train-driver. But they can’t see how to get there from where they are. Programmes which do that are very effective. The trouble is that they cost money. It looks cheaper and more effective quickly to get the young person who wants to be a chef slotted into a job vacancy as a teaching assistant.

But that is just storing up problems for the future.

I still can’t quite see mass unemployment as a result of AI. But I can see a drag on living standards and job opportunities if we become even more averse to change and fail to provide young people with the education – including higher education – they need to adjust to it.

The post David Willetts: Worrying about the effect of AI on future jobs may be distracting us from real challenges appeared first on Conservative Home.