The ro­bots might be com­ing, but a uni­ver­sal ba­sic in­come isn’t the so­lu­tion

The Globe and Mail (Prairie Edition) - - NEWS - DOUG SAUN­DERS

Giv­ing every­one a state-sup­plied $20,000 in re­sponse to ro­bots tak­ing our jobs isn’t a so­lu­tion. It’s a cop-out

Why don’t we just give every­one $20,000, no strings at­tached? That, in essence, is the pol­icy known as Uni­ver­sal Ba­sic In­come, or UBI. It has gained pop­u­lar­ity on the right and the left be­cause of its sim­plic­ity. It would re­place the ba­sic in­come al­ready pro­vided by com­plex wel­fare and unem­ploy­ment in­sur­ance plans, in­stead hand­ing an above-poverty amount to ev­ery adult re­gard­less of in­come or em­ploy­ment sta­tus.

UBI ap­peared in the news re­cently when On­tario Premier Doug Ford abruptly can­celled a pro­vin­cial pilot study that was giv­ing slightly smaller amounts to about 4,000 peo­ple. The Premier was wrong to can­cel the study; govern­ments should never be en­e­mies of knowl­edge.

But just be­cause it’s hated by small-minded peo­ple doesn’t mean UBI is a vi­able pol­icy, or even a de­sir­able one.

UBI has moved from the fringes to the main­stream be­cause some be­lieve it may soon be an emer­gency mea­sure. They be­lieve we will face mass per­ma­nent unem­ploy­ment be­cause of emerg­ing ar­ti­fi­cial in­tel­li­gence (AI) tech­nolo­gies ren­der­ing many jobs ob­so­lete, in­clud­ing a good num­ber of skilled trades and pro­fes­sions.

In­deed, a num­ber of re­cent stud­ies project that half of all cur­rently ex­ist­ing jobs will van­ish by mid-cen­tury be­cause of AI. Tax lawyers, fam­ily doc­tors, maybe den­tal hy­gien­ists will go the way of tele­phone op­er­a­tors and video­store clerks.

Un­der this logic, the only way to keep peo­ple alive – and pur­chas­ing the prod­ucts made by those ro­bots – is to give them a state-pro­vided in­come.

What­ever its in­trin­sic mer­its, a UBI pro­gram would be a po­lit­i­cal non-starter if we didn’t face hugescale job­less­ness. It would en­tail a sub­stan­tial tax in­crease on the em­ployed (off­set only partly by that free $20,000). If unem­ploy­ment re­mains in the sin­gle dig­its, it won’t hap­pen.

Yet the most in­formed and for­ward-look­ing econ­o­mists and AI ex­perts feel that even the most co­gent AI ma­chine-brains will not elim­i­nate em­ploy­ment. They may well kill half of all cur­rent jobs – but they will pro­voke the creation of at least as many as-yet-un­known jobs.

“There will al­ways be enough jobs – just the na­ture of them will change,” says Mar­cel Fratzscher, the renowned Frank­furt econ­o­mist who runs the Ger­man In­sti­tute for Eco­nomic Re­search. “There will be a thin­ning out of the mid­dle class in most Western coun­tries, be­cause those mid­dle­class jobs are the eas­i­est to re­place through tech­no­log­i­cal change, and that is al­ready hap­pen­ing.”

Branko Mi­lanovic, the City Uni­ver­sity of New York eco­nom­ics pro­fes­sor who be­came an author­ity on global in­come in­equal­ity dur­ing his years as the World Bank’s chief econ­o­mist, agrees. “I am quite op­ti­mistic on jobs un­der AI, but I’m very pes­simistic on equal­ity,” he says. “The num­ber of jobs will prob­a­bly ex­pand – but it’s an epis­te­mo­log­i­cal prob­lem, be­cause we can’t know what kind of new jobs will be cre­ated and we don’t know which ones will dis­ap­pear.”

Kai-Fu Lee, the Chi­nese AI author­ity, has a sim­i­lar un­der­stand­ing: Smart ma­chines will cre­ate at least as many jobs as they de­stroy, re­gard­less of how smart they get. And AI’s ef­fi­ciency – the only rea­son we use it – will ex­pand pro­duc­tion, and thus em­ploy­ment.

Con­sider your TV. A few decades ago, mak­ing a tele­vi­sion in­volved hun­dreds of hours of hu­man labour, so sets were once-adecade pur­chases. To­day, only a cou­ple hours labour, at most, go into mak­ing a TV. Yet far more peo­ple have jobs mak­ing TVs to­day than ever be­fore. Why? Be­cause au­toma­tion made the price of TVs plum­met to the point that most of the world can af­ford at least one ev­ery cou­ple years. It may take one-20th as many peo­ple to make a TV, but as a con­se­quence there are 40 times more TVs be­ing bought, so em­ploy­ment has at least dou­bled.

There’s no rea­son this ef­fect won’t ex­tend into ser­vices and pro­fes­sions: Ar­chi­tec­ture could go the way of TV-mak­ing – a $30,000 pro­fes­sional con­tract re­placed by a $300 in­tel­li­gent home-de­sign­ing app, with 10 times more peo­ple em­ployed as a re­sult, but at con­sid­er­ably lower in­comes.

As a re­sult, many of those econ­o­mists and AI ex­perts feel that some form of more flex­i­ble state sup­port may in­deed be needed – as part of a larger plan to pre­vent a grossly two-track so­ci­ety with half of work­ing peo­ple trapped and de­pen­dent.

But a uni­ver­sal in­come, they feel, would be not a so­lu­tion but a cop-out.

“It washes the hands of govern­ment of re­spon­si­bil­ity for the prob­lem‚” Mr. Fratzscher says. “A uni­ver­sal in­come just says ,‘ Here’ s some money, go off in a cor­ner and stop com­plain­ing.’ A job is a source of self-worth. Buy­ing peo­ple out of the labour mar­ket does not cre­ate an equal so­ci­ety, it just hides the prob­lem.”

If ma­chines get smarter, we’ll need in­tel­li­gent poli­cies – not ones that sim­ply pay peo­ple to dis­ap­pear from the hu­man com­mu­nity.

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.