Automation The Rise of the Machines Trajectory Trends

Automation: The Rise of the Machines Trajectory Trends Breakfast October 2017

Introduction

Setting the scene

What’s the problem? Employment shares and the estimated proportion of jobs at potential high risk of automation by early 2030 s for all UK industry sectors Employment share of total jobs (%) Job automation (% at potential high risk) 56% 60% 50% Will Robots Steal our Jobs? (March 17) 46% 44% 37% 40% 32% 26% 30% 17% 12% 15% 20% 26% 9% 10% 9%9% 8% 8% 24% 7% 6% 5% 4%. . n. de & n tio tra is or m in sp Pu bl ic ad an Tr fe ag or st & n tio ta & n at io od m m co Ac e n io ct on C od fo st se ac uf an M ru rv rin tu er ts or pp su & e tiv tra is in m Ad ic. . . g . io at uc Ed ifi nt ie sc l, na io ss fe Pr o Source: ONS; PIAAC; Pw. C Analysis vi. . n . . . ch c ci so & al th he an um H & al il ta re & le sa le ho W te w tra or k de 0% • PWC estimates that 30% of jobs will be at risk of automation by 2030 • More than half of these (53%) in retail, manufacturing, admin and transport

A brief history of revolutions Agricultural Industrial (second phase) Industrial (first phase) Digital

A brief history of AI 1) 1950 s: 3) 1980 s: IBM 702, the Turing Test & Dartmouth conference AI ‘winter’ – pop culture but little progress 2) 1960 -70 s: 4) 1990 s/2000 s: Advances culminate in the WABOT 1 Huge steps forward

Understanding AI 1. Deep learning – machine learning based on algorithms all of which are connected (and therefore immune to repeating mistakes) 2. Robotisation – combined with advances in robotics, AI will continue to replace human employees 3. Dematerialisation – as a result of automatic data processing/recording, autonomous software will replace many quantifiable ‘back office’ jobs 4. Gig Economy – on demand crowd working sees employees use networked platforms to access new opportunities 5. Autonomous driving – sensors allow vehicles to be self governing. Complete overhaul of transport and traffic management Source: IBA Global Employment Institute

The Pace of Change Rapid progress Barriers remain Moravec’s Paradox “…it is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility. ”

Job disruption

Some common misconceptions: #1 “Everything that can be automated will be. ” “Already today, it is technically feasible that a robotic machine could mix drinks, send the clients’ orders directly to the kitchen, receive complaints and accept the clients’ money. Nevertheless, the atmosphere in the bar or in the restaurants will no longer be the same. Because of the lack of acceptance by potential clients and the high acquisition costs, it is definite that 87 per cent of all barkeepers will not lose their jobs in the next few years. ” -- IBA Global Employment Institute

Some common misconceptions: #2 “STEM/IT skills will be the most valuable in the future. ”

Some common misconceptions: #3 “Low skilled jobs are most at risk. ” % change in employment shares by occupation group 20% 1981 -1991 15% 10% 9% 11% 1991 -2001 16% 2001 -2011 9% 5% 5% 1% 0% -5% -6% -10% -7% -11% -15% Non-routine Cognitive Source: Jaimovich/SIU, The Trend is the Cycle 2012 Routine Non-routine Manual

What jobs are safe? More Cognitive Salespeople, clerks, data entry keyers, admin and secretarial Managerial, professional occupations, analysts, doctors, economists etc High Routine Low Routine Machine operators, retail workers, assemblers, mechanics Janitors, barkeepers, gardeners, care workers More Manual

Job Polarisation: automation hollowing out the middle Measures of Recovery following Early and Recent Recessions (Jaimovich/Siu, 2012) 25 23 20 18 15 15 10 23 10 7 6 5 9 4 5 3 2 1975 1982 1991 2009 Output (half life of recession) Employment (months to turn around after recession ends) Source: Jaimovich/SIU, The Trend is the Cycle 2012 Jobless recoveries in the aggregate are accounted for by jobless recoveries in the middle-skill occupations that are disappearing. ” 90%+ of job losses during the last recession were in routine jobs 0 1970 “Job polarization is not a gradual process; essentially all of the job loss in middle-skill occupations occurs in economic downturns.

International comparisons

MAGAnomics

Choking growth Poverty Rate, India (proportion living on less than $1. 90 per day PPP) 60% 50% 40% 30% 20% 10% 0% 1983 1987 Source: World Bank (accessed 2017) 1993 2004 2009 2011

Driverless Cars The technology may arrive within 5 years, but huge barriers to adoption will remain Scope of disruption enormous: • Transport costs plunge • New era for insurance • End of courier/taxi/driver jobs

Implications

The consumer perspective All agreeing… AI will help solve complex problems in society 63% AI will help people live more fulfilling lives 59% AI will harm people by taking away jobs AI will have serious, negative implications 46% 23% 0% 10% 20% 30% 40% 50% 60% 70% Source: PWC Bot. me Report (US Adults) 80% say it’s more important to have access to more affordable legal advice than preserve the jobs of lawyers 69% Would rather have more affordable, convenient and reliable transportation than preserve the jobs of taxi drivers

A populist backlash?

Risks and rewards Automation Trajectories Take up of Automation Business efficiencies Government Tax Receipts 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Source: Trajectory

Implications Employees A new era for protest and collectivism? Business Every business is a data business? Government Data protectionism and robot taxes? Consumers Just want faster, better, cheaper?

Revolutions Create Work (1985) Ray Kurzweil looks on as Stevie Wonder experiences the Kurzweil 250, the first synthesizer to accurately reproduce the sounds of the piano — replacing piano-maker jobs but adding many more jobs for musicians “We have already eliminated all jobs several times in human history, ” said Kurzweil, pointing out that “for every job we eliminate, we’re going to create more jobs at the top of the skill ladder. … You can’t describe the new jobs, because they’re in industries and concepts that don’t exist yet. ” Why are we so bad at predicting certain things? For example, Donald Trump winning the presidency? Kurzweil: “He’s not technology. ”

November 30 th – The Future of Charities

Tom Johnson tom@trajectorypartnership. com Director Trajectory 22 Upper Ground, London, SE 1 9 PD + 44 (0)20 8004 4869 trajectorypartnership. com @Trajectory. Tweet