Monday, February 08, 2016

Race against the machine in learning ‘jobs’

Will librarians, teachers, lecturers and trainers be subject to the same pressures as other white collar jobs from tech? It's already happened, is happening and will continue to erode employment. However, it's not a simple net loss - it's a complex formula.
Economics of job losses/gains in learning game
‘Technological unemployment’ was the term coined by Keynes to describe the economic prospect of technology-driven job losses and there has been debate for centuries on the issue. Far from being a simple count of job losses against gains, it is a complex economic issue. 
Jobs in learning?
Substantial job losses will result from the effect of technology and AI. In the recent literature, however, the learning world comes out relatively unscathed. But I’m not so sure about their definition of ‘learning’. There’s a tendency for analysts and educators to see the ‘learning world’ largely through an institutional lens – schools, college, Universities. Yet this is merely a fraction of the learning world. Most learning takes place outside of these institutions, in non-institutional contexts - the workplace, at home and informally. If we define the ‘learning’ world broadly, and include all forms of cognitive improvement, we see libraries, bookshops, online bookshops, television, newspapers, online, social media and many other sources of learning, as causally relevant. When seen through this wide-angle lens, you start to realize how profound technology has been in replacing the human component in learning.
Technology and ‘learning’ jobs
Technology has always democratized, decentralized and disintermediated knowledge. It happened with the invention of writing, then the alphabet, paper and printing. These were all profound technological boosts to the transmission of knowledge. Electronically, we then had the telegraph, telephone, radio, film and TV. These ate into the time that was previously spent on print.
In the computer age, this started at the bottom of the learning pyramid, with open access to ‘knowledge’ through the internet. Google gave us access, Wikipedia and many others gave us content and jobs were lost in the publishing business (the huge encyclopedia sector died on the spot) and libraries became less useful. The fall in footfall at libraries and bookshops, I’d contend, has little to do with the fall in reading and everything to do with the rise in online access to knowledge and cheap online books. There has been a renaissance in reading and writing by young people, who read and write, it would appear, every few minutes on their phones or computers. At this level, if you want to find something out you go to the web. The one third fall in librarian jobs in the US since 1990, after decades of rises, correlates well with the rise of the computer and online access. This was a well paid, graduate-level profession that is being decimated by online technology. Similarly with large scale reference books such as encyclopedias. Wikipedia was running a worldwide publishing operation with 20/30 staff. Reference tomes and their distribution, were rapidly replaced with these online services. The effect of these services on losing knowledge delivery jobs, has been substantial. Large tech companies such as Google tend to have tens of thousands of employees, unlike previous global behemoths that had hundreds of thousands.
‘How to’ jobs also
Access to YouTube is also likely to have had an effect on practical jobs. Many, many millions of ‘how to…’ tasks have been completed by individuals who simply looked it up on YouTube. Whether it’s DIY, car repair, learning chords on a guitar… direct instruction is now available and off you go. The availability of mobile devices allows you to take the instruction to the task. Saves money, saves time – jobs lost.
Courses (organizational)
It is here that job losses are felt directly. The rise of online organizational learning has been eating away at the training market for over 30 years. This has, without doubt, led to a dramatic reduction in the number of trainers in organisations. Gone are the Stately Home training centres, and month long courses. Online learning, simulations, blended learning, flipped classroom, 70:20:10 initiatives, formal and informal, have all taken a massive, shark-size bite out of the classroom training market and undoubtedly led directly to the loss of jobs. The global corporate e-learning market will reach $51 billion in 2016 (Docebo), with healthy compound annual growth. The job losses will continue.
Courses (educational)
Again, the rise of online courses has eaten into the institutional market. It is likely to impact Higher Education, where the projected increase in human teaching staff will, to a degree, be replaced by online delivery. In the UK, the largest University in the UK is online – the Open University. Wherever online solutions perform even part of the teaching process, jobs are at risk. If you largely lecture, then that part of your job is definitely at risk, as it is easily recordable, replicable and can be distributed on scale. Global networks of Universities have already signed up to accreditation through MOOCs, more will follow. With the rise of MOOCs, where the ratio of teacher to students is huge, the reduction in teachers per student is obvious. Other trends include AI and adaptive learning that promise to eat into the teaching components, first as hybrid technology-enhanced teaching models then autonomous models. This is unlikely to have huge impact in schools, although even here, despite rising employment figures in the US, teacher jobs have been falling. Just as we have seen Uber rise but own no cars and AirBnB rise where they own no hotels, we are likely to see disintermediation on a global scale in education, where learning is delivered by organisations that may have very few teachers.
Subject shift
We may even ask what subjects are likely to rise and fall. While the teaching of English has seen rising demand, the teaching of other foreign languages in schools and at graduate level is falling. We may even be seeing a fall in liberal arts and the humanities as investment and expectations shift towards STEM subjects. Interestingly, the teaching of STEM subjects may involve more online learning and assessment, as they are less suited to ‘lecture plus essays’ delivery.
What to do?
As learning professionals we need to ask:
1) What kind of learning tasks do computers perform better than you? Low level teaching tasks such as lecturing, exposition and ‘knowledge’ transfer are most at risk, as are the handling of physical assets such as books in libraries and publishing. The textbook industry is in some trouble.
2) What kind of learning tasks do you perform better than computers? Teaching young children, practical skills, high level teaching tasks, online tutoring; these are all, as yet skills that will not be easily replaced by technology.
3) In an increasingly computerised world, what well-paid learning work is left for people to do? Online teaching, tutoring, facilitation and learning design will all be in demand, as the market for online learning grows.
4) How can people learn the skills to do this work? Get online, use social media for CPD, explore online tutoring, do a webinar, participate in MOOCs, try a MOOCs on blended learning and the use of technology in learning and learning design courses. To get some flavor of the skills needed to do online design, see here.
Economics of technology and employment
One fixture is ‘structural unemployment’, now permanent in most countries, but it is not clear that it is always caused by technology. Another are the benefits that accrue to people through technology not reflected in the job figures; cheaper prices new products for everyone. and incresed productivity. Then there are the benefits of higher wages (for those in employment), new jobs in technology, new jobs created by technology and increased investment.
While the optimistic replacement of jobs may have been true in the industrial era, many argue that it is no longer true in the information age, where fewer people design technology that replaces large numbers of blue collar, and now white collar, workers. There is a new concern, that AI can take graduate and professional jobs, just as robots took factory jobs and farm mechanization took agricultural jobs.
My own view is that we are now facing the inevitable rise of long-term structural unemployment, caused by technological innovation. Few professions will be immune to these pressures, even in the learning game. In the last couple of years economic opinion has swung in this direction. It was a major theme at Davos in 2015. However, we must be careful in not being swept away by dystopian ideas on technology, as some predict no real impact on the overall employment landscape. McKinsey, last year in 2015 published a report saying that technology tends to replace part of many jobs but not replace them. Others, like Deloittes are less optimistic. The likely outcome in many areas of education, apart from the base of the pyramid jobs such as librarians and paper publishing, is the gradual replacement of jobs by technology aids and alternatives. However, the one caveat (and it is a big caveat) is that predictions about what technology is capable of have been poor and there may be some radical and unseen shifts that come along with AI and robotics.
Postscript: Is education the cure?

An interesting brake on the problem could be the very thing we are discussing – increased delivery of education to cerate jobs. However, we must be careful in ascribing employment, and especially technological innovation and employment, to educational causes. Not all economists agree that the continual spend and expansion on Higher Education leads to increases in employment. This has been obvious in southern Europe and is now registering in graduate unemployment and underemployment figures (graduates doing non-graduate jobs). Even Paul Krugman thinks that education is not the cure. He reminds us that there has been a steady ‘hollowing out’ of middle class jobs. This is not a simple heads v hands issue, where smart people retain jobs and manual workers lose their jobs. It may be the other way round. Alan Blinder and Alan Krueger have suggested that we are now in an era where high paid jobs are easier to offshore than low paid manual jobs. Few now think that marching everyone in lock-step through an expensive higher education is the answer to technological unemployment. It may even be a placebo that leads to debt burdens that limit growth and exacerbate the problem.
Some reading….
The classic text has been Race Against the Machine (2011) by Brynjolfsson and McAfee but more recent works such as The Rise of the Robots (2015) by Martin Ford and The Future of the Professions (2015) by Susskind and Susskind are well worth a read.


No comments: