top of page
Writer's pictureGraeme Leach

The Future of AI & Work

Updated: Nov 13

An excerpt from Graeme Leach's article, 'The Future of AI and Work'.


The Future of AI & Work

 

Fears of the negative employment effects of technology have a long lineage. In 350 BC Aristotle observed that if ‘automata’ became so sophisticated they could do any work that humans did, then workers would become redundant. Aristotle’s ‘Politics’ states: “If every instrument could accomplish its own work, obeying or anticipating the will of others, like the statues of Daedalus, or the tripods of Hephaestus … the shuttle would weave and the plectrum touch the lyre without a hand to guide them, chief workmen would not want servants, nor masters slaves”.


In 1589 Elizabeth 1 refused a patent for a knitting machine, submitted by an English clergyman William Lee. In denying the patent the Queen said to Lee: "Thou aimest high Master Lee. Consider thou what the invention could do to my poor subjects. It would assuredly bring to them ruin by depriving them of employment, thus making them beggars".


In the 19th century a group of English textile workers called the Luddites – after their organiser Ned Ludd – set out to destroy textile machinery, partly because they feared powered looms were taking their jobs and livelihoods. Since then, ‘Luddite’ has become the term used to describe someone who opposes the use of new technology.


The British economist, John Maynard Keynes, worried and mused over the threat from technological unemployment. Writing in his ‘Economic Possibilities for Our Grandchildren’ (1930) he referred to the “… unemployment due to our discovery of the means of economising the use of labour outrunning the pace at which we can find new uses for labour”. Back in 1961, President John F Kennedy spoke of “… the major domestic challenge of the sixties: to maintain full employment at a time when automation is replacing men”.


But the claim now is that this time it’s different. The futurist, Jeremy Rifkin, wrote ‘The End of Work’ in 1995. In it he argued that the spread of technology was like “a deadly epidemic inexorably working its way through the marketplace, the strange, seemingly inexplicable new economic disease spreads, destroying lives and destabilising whole communities in its wake”. Rifkin promoted the idea “that within thirty years as little as 2 percent of the world’s current labour force will be needed to produce all the goods necessary for total demand”.


In 2013, Oxford academics Carl Frey and Michael Osborne, published one of the most oft quoted economic studies1 of the past decade, suggesting that 47% of US jobs (spanning 702 occupational groupings) were at high risk of being automated: “Most workers in transportation and logistics occupations, together with the bulk of office and administrative support workers and labour in production occupations, are likely to be substituted by computer capital.” Application of the Frey & Osborne methodology to the UK yielded an estimate that 35% of jobs were vulnerable to the long-term risk of automation.


Figure 1 shows that fears of the end of work have to date never been realised. If the end of work was nigh, we should be seeing it in US employment numbers. The US after all is at the leading edge of most of the new technologies. But instead of declines in aggregate US employment, we have watched it rise remorselessly over the past century.


Total US Employment (1939 - 2023)

Figure 2 approaches the same issue from a different angle, looking at the employment rate in the UK over the past 50 years. Given the advances in technology over the period, the end of work narrative would imply a declining employment rate, but the reality is no decline whatsoever. Former chief economist at the Bank of England, Andrew Haldane, commenting on the relationship between technology and employment in the UK stated: “… the historical data tell a remarkably consistent story”. Haldane cites the stability of the UK employment rate since 1750, despite wave after wave of new technology. Bank of England analysis has also shown that two-thirds of the jobs created since the Great Recession have been in full-time skilled professions. Brain work is not dead.


The 19th century philosopher and political economist John Stuart Mill captured the essential economics, stating that technological advances could “be injurious to workers” who lost jobs, but then went on to state “I do not believe that … improvements in production are often, if ever, injurious, even temporarily, to the labouring classes in the aggregate”. Mill was saying that there could be disaggregated job losses but there would be aggregate job gains. Zooming in on jobs displaced removes jobs created (replaced) from the picture.


UK Employment Rate (1971 - 2023)

Economic Theory


The fundamental theoretical issue is that the employment consequences of technology have to be seen in the full round, not just the direct impact of automation on old jobs lost. There are so called compensation effects from:


  • New technologies and the employment and investment needed to make them.

  • New investment enabled by the profits from the new technology.

  • Higher wages due to higher profitability.

  • Lower prices leading to greater demand and employment.

  • New products and processes where innovation directly creates new jobs.


The biggest changes are happening elsewhere across the whole economy, with wide ranging employment consequences from technology due to:


The Income Effect – Technological change boosts productivity growth and this in turn increases incomes and spending, boosting employment across the economy.


The Wealth Effect – The growth of new firms creates capital gains which are distributed to shareholders and employees.


The Price Effect – Productivity gains reduce costs which are then passed on to consumers in the form of lower prices. The lower prices free-up the discretionary spending power of consumers so enabling them to purchase other goods and services, which in turn creates employment elsewhere.


The displacement and replacement process can also be conceptualised in terms of direct and indirect effects. When this is done something striking emerges. The most powerful employment effects are general not specific. To focus on the jobs displaced directly by new technology and compare them with the jobs that replace them directly (producing the new technology) is to miss profound opportunities elsewhere. The first-round effect of technological innovation will be to reduce employment as man is replaced by machine. The size of this displacement effect will depend on the elasticity of substitution between labour (existing workers) and capital (the new kit). But the technology employment multiplier then kicks-in:


  • The first-round replacement effect raises employment in the new kit producer. New technology is not ex-nihilo, arising from nothing. It has to be imagined, designed, funded, built, marketed and sold. All this creates employment, the scale of which will be determined by the domestic and international market potential for the product or process.


  • The second-round replacement effect on the new kit user will depend on the price elasticity of demand for that company’s product or service. If the new kit purchased results in a significant cost saving (assuming a competitive market where cost gains are largely passed on to consumers) and price reduction in the kit user’s product or service, then an elastic demand curve could result in higher employment and/or wages in the new kit user.


  • The second-round replacement effect on the new kit producer’s sector, as a result of new entrants, creating new enterprises in the wake of the new product or process technology.


  • The second-round replacement effect on other producers is created by the price reduction for the new product or process increasing real income across the whole economy, thereby increasing the demand for goods and services generally.


  • The employment effects are then amplified through the impact on employment in the supply-chain (indirect employment) and the expenditures of those in employment (induced employment).


The total employment effect will depend on:

  1. The elasticity of substitution of labour for capital – to what extent are these factors of production substitutes or complements.

  2. The market potential for the new product or process technology and how scale-able it is.

  3. The scale of the cost saving and the elasticity of demand for the products and services of those purchasing the new product or process technology i.e. to what extend will it boost their own revenues.


In order to take a pessimistic view of future technological unemployment, the displacement effect on unemployment needs to outweigh all subsequent replacement effects on employment. This is a very big ask and history shows the positive replacement effects swamp the initial negative displacement.


In ‘The New Geography of Jobs’, Berkeley professor Enrico Moretti estimated that because of multiplier effects each new high-tech job in the US created 5 additional jobs in the service economy. Moretti writes of “the almost magical economics of job creation … for each new high-tech job in a city, five additional jobs are ultimately created outside of the high-tech sector in that city, both in skilled occupations (lawyers, teachers, nurses) and in unskilled ones (waiters, hairdressers, carpenters)”. He goes on to say that innovation, “has a disproportionate effect on the economy of American communities. Most sectors have a multiplier effect, but the innovation sector has the largest multiplier of all: about three times larger than that of manufacturing”.



This Time It's Different


Despite all the false alarms described above, the ‘Luddite Fallacy’ lives on in the 21st century. Indeed, the November 2022 launch of GPT-4 (Chat Generative Pre-Trained Transformer) from OpenAI, has spurred another wave of negative speculation regarding automation and employment. Today we have a Neo-Luddite narrative which says that ‘this time it’s different’ - just because technology has been employment positive in the past, doesn’t mean that will be in the future. In a 2015 paper, the Northwestern University economic historian, Joel Mokyr, wrote that, “if artificial intelligence and robotics continue on their present trend, future machines will be able to carry out … human capabilities at least in certain contexts and to a certain extent. Thus, it seems frighteningly plausible that this time will be different, and large sections of the labour market will be dislocated or hollowed out”.


The latest manifestation of fears over the employment consequences of new technology is driven by the rise of the robots and the potential future impact of artificial intelligence. Talk of robots and fans of Star Wars will immediately think of R2-D2 or C-3PO. For others it may trigger fears of some form of robotic apocalypse or dystopia, and the cyborg assassin played by Arnold Schwarzenegger in the film ‘Terminator’. But from an economic and practical perspective, a robot is any device or algorithm that does what humans once did. That may sound bland, but the consequences are far from it. There's immense concern that AI will do in 2 seconds what a person will do in 2 weeks, 2 months or 2 years. A Newsweek article stated that: “AI will lead us into the mother of all technological revolutions. The last time anything came close was around 1900 when the automobile, telephone, airplanes and mass electrification all came at once, radically changing the world from the late 1800s to the 1920s”.


One of the most alarmist analyses of the future impact of technology on employment and unemployment has been written by Martin Ford in ‘The Rise of the Robots – Technology and the Threat of Mass Unemployment’. Ford argues that “the robots are coming, and we have to decide now whether the future will bring prosperity or catastrophe”. Ford acknowledges that the history of technology and employment in the West over the course of the 20th century has been positive and that “advancing technology has consistently driven us towards a more prosperous society”, but he argues that the 21st century will be a very different story.


Ford argues that we face a “jobless future”. He argues that the contemporary impact of automation is much more broadly based because all business is digital nowadays in one form or another. Ford argues that most jobs can be broken down into a series of routine tasks and that we now have the capability to automate these tasks, with profound negative consequences for all routine occupations. These are tasks with well-defined procedures, which can be replicated by sophisticated algorithms. But this is more than just automation, it is the capacity for machines to talk to and learn from each other. According to Ford, it is the routine characteristic which determines the vulnerability to automation, whether it is in routine cognitive or routine manual tasks. And the reasoning behind the view that technology will eat its way up the occupational ladder, is examples of hitherto skilled/highly skilled tasks already being taken on by a machine, and the perceived future potential for these technologies.


One possible justification for the ‘this time it’s different’ story is that AI hasn’t been widely employed up until now. In other words, the robots are still coming for our jobs but they’re a bit slow getting there. But a 2022 business survey by McKinsey found that 50% of respondents had adopted AI in at least one business unit or function, up from 20% in 2017. Granted AI has huge future potential, and we’re only at the bottom of the S curve thus far, but even so, there is still no sign of any negative employment effects in the data. It is claimed that significant effects are still around the corner though.


A major study by PWC for HM Government estimated that up to 30% of UK jobs could be at risk of automation over the next 20 years, but that only 20% will eventually be automated. Over the 2017-2037 period PWC estimate that in the UK around 7 million jobs will be displaced by AI and around 7.2 million will be created.


PWC estimate that replacement will match displacement over the two decades to 2037. But this neutral assessment does not mean that the tectonic plates of employment will hardly move. PWC argue that automation will unfold in 3 very significant waves:


An Algorithm Wave – the automation of simple computational tasks, which is well underway.


An Augmentation Wave – the automation of repeatable tasks, this is also underway and likely to mature in the 2020s.


An Autonomy Wave – the automation of physical labour, manual dexterity and problem solving in dynamic real-world situations, likely to mature in the 2030s.


So, we can say with some certainty that there is a great deal of uncertainty as to what the eventual employment consequences of AI will be. Proponents of ‘this time it’s different’ have far from given up. The uncertainty has been increased by additional theories.


In ‘The Globotics Upheaval – Globalization, Robotics and the Future of Work’ (2019), Richard Baldwin, professor of economics at the Graduate Institute Geneva, argues that we are entering an era of both a new form of globalisation and a new form of robotics. He combines the two to call it ‘globotics’.


Baldwin argues that the threat to jobs from AI will be compounded by the threat from remote intelligence (RI). Baldwin describes RI as telemigration, or telecommuting gone global. Baldwin argues that there is an enormous shift underway, with highly skilled, but low-cost foreign workers, ushering in a new era of globalisation, and competing with highly skilled service sector workers in the advanced economies. Facilitated by big strides in machine translation, Baldwin foresees an RI ‘talent tsunami’. This is only half the thesis though. Baldwin also highlights the future impact of white-collar robots and machine learning. The end result, according to Baldwin, is that: “RI and AI are coming for the same jobs at the same time and driven by the same technologies”.


Baldwin states: “Automation and globalisation are century-old stories. Globotics is different for two big reasons. It is coming inhumanly fast, and it will seem unbelievably unfair”. He argues that this time really will be different with a “radical mismatch between the speed of jobs displacement and the speed of job replacement”. The consequences, according to Baldwin, will be shocking: “Globotics is injecting pressure into our socio-political system (via job displacement) faster than our system can absorb it (via job replacement). This may break the societal confinements that restrain hostility and violent reactions. The result could be blast waves that travel considerable distances before they dissipate. Deep down, the explosive potential comes from the mismatch between the speed at which disruptive energy is injected into the system by job displacement and the system’s ability to absorb it with job creation. The displacement is driven at the eruptive pace of digital technology; the replacement is driven by human ingenuity which moves at the leisurely pace it always has”.



No, It's Not Different


We’ll now examine why there’s nothing new under the sun; starting by going back to the future. James Bessen, an economist at Boston University School of Law, was quoted in The Economist as saying that: “During the Industrial revolution more and more tasks in the weaving process were automated, prompting workers to focus on the things machines could not do, such as operating a machine and then keeping multiple machines operating smoothly. This caused output to grow explosively. In America during the 19th century the amount of coarse cloth a single weaver could produce in an hour increased by a factor of 50, and the amount of labour required per unit of cloth fell by 98 percent. This made cloth cheaper and increased demand for it, which in turn created more jobs for weavers: their numbers quadrupled between 1830 and 1900. In other words, technology gradually changed the nature of the weaver’s job, and the skills required to do it, rather than replacing it altogether”. Commenting on this, The Economist noted, “… rather than destroying jobs automation redefines them, and in ways that reduce costs and increase demand”. By the middle of the 19th century Britain produced half the world’s cotton, because it had embraced the new technology of the time.


A study by Deloitte on the relationship between jobs and the rise of technology, trawled through Census data for England and Wales, going back to 1871. Their conclusion was unremittingly positive: “The last 200 years demonstrates that when a machine replaces a human, the result, paradoxically, is faster growth and, in time, rising employment ... UK employment has doubled in the past 150 years”.


Fast forward to recent times, and Bessen’s research shows that over the 30-year period 1982-2012, employment grew significantly faster (1.7 times) in occupations that made more use of computers. According to Bessen, the net effect was that more computer intensive jobs within an industry displaced less computer intensive ones. His research suggests that computers reallocate rather than displace jobs.


A report from the Brookings Institute in the US highlights the economic transmission mechanism which leads to more not less jobs: “The fear that automation will eliminate millions of jobs, leaving masses of workers jobless, has periodically emerged in industrialized countries at least since the Luddites first made that claim in Britain in the mid-19th century. In the US, such fears occasionally surface as well, as they did during a brief ‘automation scare’ in the late 1950’s and early 1960’s, when a wide swathe of workers felt some risk of displacement. To date, these fears have never proven accurate in any industrial country. New jobs always emerge to replace those that have been lost. This is true because automation raises worker productivity and reduces the costs and prices of goods and services, which makes consumers richer. They can now afford to buy more products than before, which then creates new jobs for workers to fill”.


David Autor, an economist at MIT, argues that there is “zero evidence” that AI is having a new and significantly different impact on employment. Autor counsels caution to the pessimists: “… just as people worry about the impact of self-driving cars today, a century ago there was much concern about the impact of the switch from horses to cars. Horse related jobs declined, but entirely new jobs were created in the motel and fast-food industries that arose to serve motorists and truck drivers”. Autor argues that “… in the past technology has always ended up creating more jobs than it destroys. That is because of the way automation works in practice … automating a particular task so that it can be done more quickly or cheaply, increases the demand for human workers to do the other tasks around it that haven’t been automated”.


The Information Technology & Innovation Foundation (ITIF) in the United States, have quantified a ratio of technology creating new jobs to technology eliminating old jobs and state that: “One of the prevailing narratives about the US labour market is that technology is destroying more jobs than it is creating. But that has been true for the past 165 years”. This is worth repeating and restating for emphasis. Technology always destroys more old jobs (directly) than it creates new jobs (directly). The bigger story, according to the ITIF, is that “most new jobs will not be created in the new machinery firms. Rather they will be created across the economy from the new demand that higher productivity enables”.


Total employment statistics don’t tell the whole story. Trend growth in total employment might hide an enormous amount of ‘churning’ within it, with considerable disruption from exit and entry into employment. However, when you analyse the churning effect something rather surprising emerges. The ITIF state: “When we actually examine the last 165 years of American history, statistics show that the US labour market is not experiencing particularly high levels of labour churn (defined as the sum of the absolute values of jobs added in growing occupations and jobs lost in declining occupations). In fact, it’s the exact opposite: Levels of occupational churn in the United States are now at historic lows … contrary to popular perception … the rate of occupational churn in recent decades is at the lowest level in American history”.


ITIF state that “The levels of churn in the last 20 years … purported to be more powerfully disruptive than anything in the past – have been just 38 percent of the levels from 1950 to 2000, and 42 percent of the levels from 1850 to 2000”. ITIF conclude that: “The single biggest economic challenge facing advanced economies today is not too much labor market churn, but too little, and thus too little productivity growth … the bigger risk is that economies will not be able to raise productivity fast enough to adequately raise per-capita incomes, especially in an era where nations face growing elderly populations”. As a result, the ITIF say that tech-driven labour market disruption is a myth.



Concluding Thoughts


As a general rule, those most optimistic about the potential from new technologies such as AI, tend to be least optimistic about the employment consequences. But Kingsgate is not one of them. Kingsgate argues that there is an optimistic case for both technology and the employment consequences. Moreover, we shouldn’t be surprised at the current angst. The eminent economic historian Joel Mokyr has written that: “From generation to generation, literature has often portrayed technology as alien, incomprehensible, increasingly powerful and threatening, and possibly uncontrollable”. Societies that have established norms and habits to fit the previous technological revolution struggle to assimilate the new.


Jeremy Warner has written in The Daily Telegraph that: “All technological revolutions will in time create far more jobs than they destroy, but they often involve long, difficult, and socially disruptive transitions”.


One final consideration also is how technology could reverse manufacturing employment decline in the West, dampening or even reversing the shift in the geopolitical centre of gravity from West to East. This could arise if 3D additive manufacturing leads to the substantial reshoring of manufacturing activity. This is a topic we will return to in a future paper.



Full Read:

Click / tap below to download, read and use Graeme Leach’s article ‘The Future of AI and Work’ in full now.


 

Kingsgate is a professional services firm committed to the growth and transformation of your organization. We are dedicated to helping you unlock enduring value, even in volatile seasons, and bring transformation that lasts.


To learn more about our Transaction Advisory and Growth Advisory consulting services, then please complete the Contact Us form.


See other articles from Kingsgate:








Comentarios


bottom of page