Are You Ready for Ro­bots?

Trillions - - Contents - by Martin Khor

Last year Uber started test­ing driver-less cars, with hu­mans in­side to make cor­rec­tions in case some­thing goes wrong. If the tests go well, Uber will pre­sum­ably re­place their present army of driv­ers with fleets of the new cars. Some per­son­ally owned cars can al­ready do au­to­matic park­ing. Is it a mat­ter of time be­fore Uber, taxi and per­sonal ve­hi­cles will all be smart enough to bring us from A to B with­out our having to do any­thing our­selves? But in this ap­pli­ca­tion of "ar­ti­fi­cial in­tel­li­gence", in which ma­chines can have hu­man cog­ni­tive func­tions built into them, what will hap­pen to jobs? It is es­ti­mated that in the US alone, 4 to 5 mil­lion driv­ers of trucks and taxis could be ren­dered un­em­ployed. The driver-less ve­hi­cle is just one ex­am­ple of the tech­no­log­i­cal rev­o­lu­tion that is go­ing to dras­ti­cally trans­form the world of work and liv­ing. There is con­cern that the march of au­to­ma­tion tied with dig­i­tal tech­nol­ogy will cause dis­lo­ca­tion in many fac­to­ries and of­fices, and even­tu­ally lead to mass un­em­ploy­ment. Just a day be­fore he left of­fice, Barack Obama warned in a farewell in­ter­view that "jobs are go­ing away be­cause of au­to­ma­tion and that's go­ing to ac­cel­er­ate," point­ing to "driver­less Uber" and "dis­place­ment that's go­ing to take place in of­fice build­ings across the coun­try." Also voic­ing con­cern about the so­cial im­pact of au­to­ma­tion, Mi­crosoft founder Bill Gates re­cently pro­posed that gov­ern­ments should im­pose a tax on ro­bots. Com­pa­nies us­ing ro­bots should have to pay taxes on the in­comes at­trib­uted to the use of ro­bot­ics. That pro­posal has caused an up­roar, with main­stream econ­o­mists like Lawrence Sum­mers, a for­mer US trea­sury sec­re­tary, con­demn­ing it for putting brakes on tech­no­log­i­cal ad­vance­ment. One critic sug­gested that the first com­pany to pay taxes for caus­ing au­to­ma­tion should be Mi­crosoft. How­ever, the tax on ro­bots idea is one re­sponse to grow­ing fears that the au­to­ma­tion rev­o­lu­tion will in­crease in­equal­ity as many lose their jobs while a few reap the ben­e­fits of in­creased pro­duc­tiv­ity and prof­itabil­ity. The new technologies will cause un­con­trol­lable dis­rup­tion and add to the so­cial dis­con­tent and po­lit­i­cal up­heaval in the West which had fu­elled the anti-es­tab­lish­ment votes for Brexit and Don­ald Trump. Re­cent stud­ies are show­ing that deep­en­ing use of au­to­ma­tion will cause wide­spread dis­rup­tion in many sec­tors and even whole economies. Worse, it is the de­vel­op­ing coun­tries that are es­ti­mated to lose the most, and this will ex­ac­er­bate the al­ready great global in­equal­i­ties.

The risk of au­to­ma­tion to jobs in de­vel­op­ing coun­tries is es­ti­mated to range from 55 to 85 per cent, ac­cord­ing to a study in 2016 by Ox­ford Univer­sity's Martin School and Citi. Ma­jor emerg­ing economies will be at high risk, in­clud­ing China (77%) and In­dia (69%), higher than the OECD de­vel­oped coun­tries' av­er­age risk of 57%. The Ox­ford-citi re­port, "The fu­ture is not what it used to be", pro­vides many rea­sons why the au­to­ma­tion rev­o­lu­tion will be par­tic­u­larly dis­rup­tive in the de­vel­op­ing coun­tries. First, there is "pre­ma­ture dein­dus­tri­al­i­sa­tion" tak­ing place as man­u­fac­tur­ing is be­com­ing less labour-in­ten­sive and many de­vel­op­ing coun­tries have reached the peak of their man­u­fac­tur­ing jobs. Man­u­fac­tur­ing pro­cesses are more au­to­mated to­day, also in low and mid­dle in­come de­vel­op­ing coun­tries. Sec­ond, while 20th cen­tury technologies al­lowed com­pa­nies to shift pro­duc­tion abroad to take ad­van­tage of cheap labour, re­cent de­vel­op­ments in ro­bot­ics and ad­di­tive man­u­fac­tur­ing now en­able firms to lo­cate pro­duc­tion closer to do­mes­tic mar­kets in au­to­mated fac­to­ries. Sev­enty per cent of clients sur­veyed be­lieve au­to­ma­tion and 3D print­ing de­vel­op­ments will en­cour­age com­pa­nies to move their man­u­fac­tur­ing close to home. China, ASEAN and Latin Amer­ica have the most to lose from this re­lo­ca­tion, while North Amer­ica, Europe and Ja­pan are the main win­ners. Thirdly, "the im­pact of au­to­ma­tion may be more dis­rup­tive for de­vel­op­ing coun­tries, due to lower lev­els of con­sumer de­mand and limited so­cial safety nets" as com­pared to the de­vel­oped coun­tries, ac­cord­ing to a sum­mary of the Ox­ford Martin School re­port. The re­port warns that de­vel­op­ing coun­tries may even have to re­think their over­all de­vel­op­ment mod­els as the old ones that were suc­cess­ful in gen­er­at­ing growth in the past will not work any­more. "In the light of these tech­no­log­i­cal de­vel­op­ments, in­dus­tri­al­iza­tion is likely to yield sub­stan­tially less man­u­fac­tur­ing em­ploy­ment in the next gen­er­a­tion of emerg­ing economies than in the coun­tries pre­ced­ing them. Hence it will be in­creas­ingly dif­fi­cult for African and South Amer­i­can man­u­fac­tur­ing firms to cre­ate jobs in the same num­bers that Asian coun­tries have done. In other words, to­day's low-in­come coun­tries will not have the same pos­si­bil­ity of achiev­ing rapid growth by shift­ing work­ers from farms to higher-pay­ing fac­tory jobs." In­stead of ex­port-led man­u­fac­tur­ing growth, de­vel­op­ing coun­tries will need to search for new growth mod­els, said the re­port. "Ser­vice-led growth con­sti­tutes one op­tion, but many low-skill ser­vices are now be­com­ing equally au­tomat­able." It cites a World Bank re­port show­ing de­vel­op­ing coun­tries are highly sus­cep­ti­ble to their work­force be­ing af­fected by in­creas­ing au­to­ma­tion, even rel­a­tive to ad­vanced economies where labour costs are high. More­over, coun­tries with lower lev­els of GDP per capita typ­i­cally have a higher share of their work­force "at risk". "Thus there are rea­sons to be con­cerned about the fu­ture of in­come con­ver­gence, as low in­come coun­tries are rel­a­tively vul­ner­a­ble to au­to­ma­tion," con­cludes the re­port. An­other se­ries of re­ports, by Mckin­sey Global In­sti­tute, found that 49% of present work ac­tiv­i­ties can be au­to­mated with cur­rently demon­strated tech­nol­ogy, and this trans­lates into US$15.8 tril­lion in wages and 1.1 bil­lion jobs glob­ally. About 60% of all oc­cu­pa­tions could see 30% or more of their ac­tiv­i­ties au­to­mated and 5% of jobs can be en­tirely au­to­mated. But more re­as­sur­ingly an au­thor of the re­port James Manyika says the changes will take decades. How au­to­ma­tion af­fects jobs will not be de­cided sim­ply by what is tech­ni­cally fea­si­ble. Other fac­tors in­clude eco­nomics, labour mar­kets, reg­u­la­tions and so­cial at­ti­tudes. Which jobs are most sus­cep­ti­ble to be af­fected? While most peo­ple think they would be in man­u­fac­tur­ing, in fact many jobs in ser­vices will also be dis­rupted. The Mckin­sey study lists ac­com­mo­da­tions and food ser­vices as the most vul­ner­a­ble sec­tor in the US, fol­lowed by man­u­fac­tur­ing and re­tail busi­ness. In ac­com­mo­da­tions and food, 73% of ac­tiv­i­ties work­ers per­form can be au­to­mated, in­clud­ing pre­par­ing, cook­ing or serv­ing food; clean­ing food-prepa­ra­tion ar­eas, pre­par­ing bev­er­ages and col­lect­ing dirty dishes. In man­u­fac­tur­ing, 59% of all ac­tiv­i­ties can be au­to­mated, es­pe­cially physical ac­tiv­i­ties or op­er­at­ing ma­chin­ery in a pre­dictable en­vi­ron­ment. Ac­tiv­i­ties range from pack­ag­ing prod­ucts to load­ing ma­te­ri­als on pro­duc­tion equip­ment to weld­ing to main­tain­ing equip­ment. For re­tail­ing, 53% of ac­tiv­i­ties are au­tomat­able. They in­clude stock man­age­ment, pack­ing ob­jects, main­tain­ing sales records, gath­er­ing cus­tomer and prod­uct in­for­ma­tion, and ac­count­ing. A tech­nol­ogy spe­cial­ist writer and con­sul­tant, Shelly Palmer, has also listed elite white-col­lar jobs that are at risk from "ro­bots" which she de­fines as technologies, such as ma­chine learn­ing al­go­rithms run­ning on pur­pose-built com­puter plat­forms, that have been trained to per­form tasks that cur­rently re­quire hu­mans to per­form. Those she as­sessed would be dis­placed in­clude mid­dle man­agers, sales­per­sons, re­port writ­ers, jour­nal­ists and an­nounc­ers, ac­coun­tants, book­keep­ers and doc­tors.

While some an­a­lysts are en­thu­si­as­tic about the pos­i­tive ef­fects of the au­to­ma­tion rev­o­lu­tion, oth­ers are alarmed by its ad­verse ef­fects. Cer­tainly, the tech­no­log­i­cal trend will im­prove pro­duc­tiv­ity per worker that re­mains, and in­crease the prof­itabil­ity of com­pa­nies that sur­vive. While there are ben­e­fits at the mi­cro level for those com­pa­nies and in­di­vid­u­als that thrive in the new en­vi­ron­ment, there are ad­verse ef­fects at macro level, es­pe­cially re­trench­ment for those whose jobs are no longer needed. What can be done to slow down au­to­ma­tion or at least to cope with its ad­verse ef­fects? The Bill Gates pro­posal to tax ro­bots is one of the most rad­i­cal. The tax could slow down the tech­no­log­i­cal changes and the funds gen­er­ated by the tax could be used to mit­i­gate the so­cial ef­fects. An­other rad­i­cal idea which is gen­er­at­ing a lot of de­bate is to pro­vide "univer­sal in­come" to ev­ery­one ir­re­spec­tive of whether they are work­ing. The high pro­duc­tiv­ity will al­low ev­ery­body to be paid a com­fort­able in­come, and thus there is no need to worry that au­to­ma­tion will dis­place jobs. Gov­ern­ments can also take the at­ti­tude of "join them if you can't beat them." For ex­am­ple, China is see­ing ma­jor op­por­tu­ni­ties in join­ing the tech­no­log­i­cal rev­o­lu­tion and has drawn up plans to in­vest in ro­bot­ics and ar­ti­fi­cial in­tel­li­gence. Other more con­ven­tional pro­pos­als in­clude up­grad­ing the ed­u­ca­tion of stu­dents and present em­ploy­ees to take on the new jobs re­quired in man­ag­ing or work­ing with the au­to­mated pro­duc­tion process, and train­ing work­ers to be made redundant with the new skills needed to work in the new en­vi­ron­ment. Over­all, how­ever, there is likely to be a net loss of em­ploy­ment, at least in the short term, and thus the po­ten­tial for so­cial dis­con­tent. As for the de­vel­op­ing coun­tries in gen­eral, there will have to be much think­ing of the im­pli­ca­tions of the new technologies for their im­me­di­ate and long-term eco­nomic prospects, and a ma­jor re­think­ing of eco­nomic and de­vel­op­ment strate­gies is also called for.

Im­age by denisgo

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.