Be­ing hu­man: how re­al­is­tic do we want ro­bots to be?

With Google’s AI as­sis­tant able to make phone calls and an­droids pop­u­lat­ing house­holds in games and films, the line be­tween ma­chine and man is get­ting scar­ily blurred

The Myanmar Times - Weekend - - Weekend | Tech -

A S our de­pen­dence on tech­nol­ogy builds and the pri­vacy-destroying, brain­hack­ing con­se­quences of that start to come to light, we are see­ing the re­turn of a sci­ence-fic­tion trope: the rise of the ro­bots. A new wave of tele­vi­sion shows, films and video games is grap­pling with the ques­tion of what will hap­pen if we de­velop the tech­nol­ogy to cre­ate machines in our own im­age.

West­world posits that if we could de­velop re­al­is­tic an­droids, we would want to rape and mur­der them for fun. In Blade Run­ner 2049, they have re­placed hu­mans as sex work­ers and man­ual labour­ers. In the re­cently re­leased video game Detroit: Be­come Hu­man, an­droids are nan­nies, car­ers and even pop stars, om­nipresent in the home and in city life.

The cur­rent wave of an­droid fic­tion cen­tres on what hap­pens when the line be­tween hu­man and ma­chine be­comes blurred. At what point do ro­bots de­serve rights: when they reach a cer­tain level of in­tel­li­gence, or when they de­velop the ca­pac­ity for emo­tion, cre­ativ­ity or free will? In the cold war, when we be­lieved that machines might kill us any minute in the shape of nu­clear bombs, our night­mare ro­bots were re­lent­less killing machines such as The Ter­mi­na­tor or Robo­cop – or the piti­less mil­i­tary droids that hunt down the last rem­nants of hu­man­ity in Me­tal­head, a re­cent episode of Black Mir­ror. Now that tech­nol­ogy has en­meshed it­self in our lives, it is dawn­ing on us that machines can take over in an­other way – by en­croach­ing on our hu­man­ity.

Just a few weeks ago, Google demon­strated that its home­as­sis­tant ro­bot is ca­pa­ble of hold­ing an un­set­tlingly nat­u­ral con­ver­sa­tion with a hu­man be­ing over the phone to book a hair­cut or make a restau­rant reser­va­tion, com­plete with “ums” and “ahs” to make the lis­tener be­lieve they are talk­ing to a real per­son.

We are in­creas­ingly wor­ried about what will hap­pen if machines be­come just like us. Adam Wil­liams, lead writer on the game Detroit: Be­come Hu­man, thinks that the devel­op­ment of hu­man-like emo­tion is more un­set­tling than the idea of straight­for­ward ro­bot an­tag­o­nism. “It’s a more sub­tle threat to the sanc­tity of the hu­man cat­e­gory,” he says. “Emo­tion is some­thing we re­serve for our­selves: depth of feel­ing is what we use to jus­tify the pri­macy of hu­man life. If a ma­chine is ca­pa­ble of feel­ing, that doesn’t make it dan­ger­ous in a Ter­mi­na­tor-es­que fash­ion, but in the ab­stract sense of im­ping­ing on what we think of as clas­si­cally hu­man.”

In the game, house­hold an­droids that have been mis­treated by hu­mans start re­belling, even­tu­ally band­ing to­gether to de­mand rights. It is not an orig­i­nal premise, but video games now look so life­like that it is a good lit­mus test for how com­fort­able you feel with the idea of a hu­man-like an­droid. The game’s char­ac­ters, played by hu­man ac­tors, look al­most in­dis­tin­guish­ably close to real peo­ple.

Anouk van Maris, a ro­bot cog­ni­tion spe­cial­ist who is re­search­ing eth­i­cal hu­man-ro­bot in­ter­ac­tion, has found that com­fort lev­els with ro­bots vary greatly de­pend­ing on lo­ca­tion and cul­ture. “It de­pends on what you ex­pect from it. Some peo­ple love it, oth­ers want to run away as soon as it starts mov­ing,” she says. “The ad­van­tage of a ro­bot that looks hu­man-like is that peo­ple feel more com­fort­able with it be­ing close to them, and it is eas­ier to com­mu­ni­cate with it. The big dis­ad­van­tage is that you ex­pect it to be able to do hu­man things and it of­ten can’t.”

In Ja­pan, where the an­i­mus be­lief per­haps makes peo­ple more com­fort­able with the idea that spirit can re­side in some­thing that isn’t hu­man, ro­bots are al­ready be­ing used as shop as­sis­tants, in care homes and in schools. Ja­pan is the world leader in ro­bot­ics and de­mand is high for ro­bots that could help fill a short­fall in nurs­ing care. The coun­try is home to the creepy Erica, the most re­al­is­tic fe­male hu­manoid in ex­is­tence, and Gate­box AI’S Azuma, a holo­graphic girl in a jar that com­bines Alexa-like home-as­sis­tant func­tion­al­ity with a cute anime look and a sim­u­lated, def­er­en­tial per­son­al­ity.

In Europe, by con­trast, peo­ple are gen­er­ally un­com­fort­able with the idea of an an­droid per­form­ing roles that re­quire in­ter­ac­tion with hu­mans. “In one study, peo­ple were asked if there was a ro­bot in­ter­act­ing with chil­dren, whether it would be eth­i­cally ac­cept­able if the chil­dren got at­tached to that ro­bot,” says Van Maris. “Only 40% thought that was ac­cept­able.” It is telling that US com­pa­nies de­sign their home-as­sis­tant ro­bots to look like black boxes and sound like com­put­ers.

“A ma­chine can ex­hibit hu­man­like qual­i­ties and not be con­sid­ered par­tic­u­larly con­tro­ver­sial if it doesn’t look hu­man,” says Wil­liams. “That’s what is in­trigu­ing. What scares peo­ple about that Google As­sis­tant phone call is that it sounds hu­man. The fact that it can con­struct the con­ver­sa­tion is not what scares peo­ple – it’s the fact they can’t dis­tin­guish it from a real per­son.”

Some ro­bot­ics ex­perts, in­clud­ing the Univer­sity of Ed­in­burgh’s Robert Fisher, see the con­cept of hu­man-like ro­bots as ill-ad­vised. “I don’t think ar­ti­fi­cial in­tel­li­gence will ever be like hu­mans,” Fisher says. “We put our­selves and them in a dif­fi­cult sit­u­a­tion by try­ing to pre­tend they are hu­man, or make them look like us. Maybe it is bet­ter not to do that in the first place. Sex ro­bots is per­haps the only case where there is a rea­son for them to look hu­man.”

On the ev­i­dence of West­world, Detroit: Be­come Hu­man and Ex Machina – none of which paint the most op­ti­mistic por­trait of hu­man an­droid re­la­tions – per­haps we will all be bet­ter off if our fu­ture ro­bot as­sis­tants are more Wall-e or R2-D2 in Star Wars than Star Trek’s Data or Blade Run­ner’s Pris.

Photo: the Guardian

Om­nipresent … in Detroit: Be­come Hu­man an­droids are nan­nies, car­ers and pop stars.

Newspapers in English

Newspapers from Myanmar

© PressReader. All rights reserved.