Shout-out to sus­pi­cion and sys­tem er­rors

Over­all, the year was a bit of a damp squib for the tech sec­tor, notes Rho­dri Mars­den

The National - News - - ARTS & LIFESTYLE -

Com­put­ers made strides for­ward in ei­ther guess­ing what we wanted, or do­ing things on our be­half

Tra­di­tion­ally, an end-of-year re­cap of 12 months in the world of tech­nol­ogy tends to cel­e­brate hu­man in­ge­nu­ity and all its pos­i­tive ef­fects. This, how­ever, was not a feel-good year; new giz­mos, apps and ser­vices tended to be greeted by doubts as to whether they were re­ally for our ben­e­fit, or merely for the ben­e­fit of the com­pa­nies who pro­duced them. That kind of cyn­i­cism has al­ways ex­isted among in­dus­try crit­ics, but this year saw sus­pi­cion and con­cern among peo­ple who would nor­mally feel en­thused by tech­no­log­i­cal in­no­va­tion.

That isn’t to say we stopped spend­ing money. Wear­able tech­nol­ogy – par­tic­u­larly smart­watches – may have fallen short of mass mar­ket ap­peal once again, but pub­lic en­thu­si­asm for voice-pow­ered as­sis­tants such as Ama­zon Echo and Google Home con­tin­ued to grow. Part of that was down to savvy mar­ket­ing, but there were tech­no­log­i­cal im­prove­ments, too, Mi­crosoft claimed this year that its voice-recog­ni­tion sys­tem now has an er­ror rate of just 5.1 per cent – com­pa­ra­ble to hu­man tran­scrip­tion – while as­sis­tants have sim­ply be­came bet­ter at as­sist­ing. Fond­ness for Ap­ple’s Siri, Ama­zon’s Alexa and Mi­crosoft’s Cor­tana grew as deep learn­ing tech­niques re­sulted in fewer er­rors and re­duced frus­tra­tion.

But the genre was by no means flaw­less. Sam­sung’s voice as­sis­tant, Bixby, was poorly re­ceived, while Ama­zon’s Echo Look, a “hands-free cam­era and style as­sis­tant” of­fer­ing fash­ion ad­vice, prompted puz­zle­ment best summed up by Ars Tech­nica’s Jeff Dunn. “The idea of pay­ing $200 [Dh734.5] to be judged on your ap­pear­ance by a semi-in­tel­li­gent as­sis­tant with a mas­sive stake in get­ting you to buy new out­fits is just non­sen­si­cal,” he wrote.

All around us, com­put­ers made strides for­ward in ei­ther guess­ing what we wanted, or do­ing things on our be­half. The most strik­ing man­i­fes­ta­tion of this was in Phoenix, Ari­zona, where Google’s self-driv­ing car project, Waymo, be­gan test­ing driver­less cars with­out any hu­man su­per­vi­sion. As these kinds of tech­no­log­i­cal leaps were made, voices of cau­tion would reg­u­larly sound; not just in terms of hu­man safety, but also the wider is­sue of de­ci­sions be­ing taken by com­put­ers and the ra­tio­nale be­hind the al­go­rithms that power them. We must not, crit­ics said, ac­cept al­go­rithms as in­her­ently wise. “[They] are not neu­tral,” wrote En­gad­get’s Chris Ip, “and the rea­son they’re bi­ased is that so­ci­ety is bi­ased.” The de­bate over whether ar­ti­fi­cial in­tel­li­gence (AI) will have our best in­ter­ests at heart, be­came heated in late July as the out­spo­ken CEO of SpaceX, Elon Musk, pub­licly voiced his wor­ries over “the risk … to hu­man civil­i­sa­tion” – but Face­book’s Mark Zucker­berg dis­missed this, as­sur­ing the pub­lic that for his part, at least, AI was about “mak­ing peo­ple’s lives bet­ter”.

Face­book, how­ever, would strug­gle to prove it­self as a whole­some ben­e­fit to so­ci­ety in 2017, and Google and Twit­ter didn’t fare much bet­ter. All three firms were sum­moned by the United States Congress to tes­tify over Rus­sian in­flu­ence in the US pres­i­den­tial elec­tion last year, and all three ad­mit­ted that Rus­sian-bought ads had at­tempted to af­fect the vote.

Pol­i­tics was just one strand of con­cern, along­side the pro­mo­tion of fake news sto­ries, in­ap­pro­pri­ate chil­dren’s en­ter­tain­ment, the pro­lif­er­a­tion of on­line ha­rass­ment and many oth­ers. Blame for this was lev­elled at the cul­ture of Sil­i­con Val­ley, from its hell-bent com­mit­ment to growth at any cost, to its notable lack of di­ver­sity. An il­lus­tra­tion of the strength of feel­ing sur­round­ing this is­sue came in the form of an es­say by Google en­gi­neer, James Damore, who ques­tioned the value of di­ver­sity in the work­place at great length. He was dis­missed by Google, trans­form­ing him into both a fig­ure of de­ri­sion and a cause célèbre overnight.

Away from the con­tro­versy, some notable prod­ucts were mak­ing their way into stores. The Sam­sung Galaxy S8 and the iPhone X were lauded for their de­sign and their ca­pa­bil­i­ties; mul­ti­ple sen­sors and pow­er­ful cam­eras trans­formed these high-end smart­phones into po­tent “aug­mented re­al­ity” de­vices. The much-dis­cussed “notch” at the top of the iPhone X was home to an ar­ray of sen­sors that could, ac­cord­ing to Ap­ple, en­able the de­liv­ery of new ser­vices and new en­ter­tain­ment; ex­am­ples in­cluded see­ing what par­tic­u­lar items of fur­ni­ture might look like in your liv­ing room, or plac­ing friends and fic­tional char­ac­ters in the same vir­tual space. That notch also fa­cil­i­tated Ap­ple’s new fa­cial­recog­ni­tion sys­tem, Face ID, prompt­ing de­bate over the safety and se­cu­rity of fa­cial recog­ni­tion in a con­sumer prod­uct.

In this par­tic­u­lar case, Ap­ple was able to re­as­sure the pub­lic that it col­lected no fa­cial im­age data, but con­cern over per­sonal data col­lec­tion now hov­ers over ev­ery new, cloud-con­nected prod­uct. The ques­tion of how much data com­pa­nies pos­sess and how they store it, is raised ev­ery time a se­cu­rity breach hits the news; well-pub­li­cised cases in 2017 in­cluded Equifax, Ya­hoo and Uber (which had a stink­ing year for many rea­sons, from the po­ten­tial loss of its Lon­don li­cence to the res­ig­na­tion of its co-founder, Travis Kalan­ick).

These con­cerns over the vul­ner­a­bil­ity of data prompted in­ter­est in the amor­phous con­cept of the blockchain, a dig­i­tally ver­i­fied ledger that has found its most fa­mous use in the cryp­tocur­rency Bit­coin. The in­abil­ity of most of us to un­der­stand how the Bit­coin works didn’t stop its value from rock­et­ing to­wards the end of the year, as in­vest­ments were si­mul­ta­ne­ously en­cour­aged and cau­tioned against.

Bit­coin has en­gen­dered a feel­ing among many of its users, of them not know­ing ex­actly what they’re play­ing with, but that sen­sa­tion ex­tends across many tech­no­log­i­cal de­vel­op­ments this year. As com­plex ques­tions are raised about the ef­fect of tech­nol­ogy on our fu­ture, the sub­se­quent de­bates tend ei­ther to be im­prop­erly con­ducted or sav­agely cur­tailed; this cer­tainly hap­pened ear­lier this month, when the US Fed­eral Com­mu­ni­ca­tions Com­mis­sion de­cided to dis­man­tle its rules over net neu­tral­ity, paving the way for US in­ter­net ser­vice providers to pri­ori­tise or block par­tic­u­lar apps or web­sites. This, cam­paign­ers say, will have a chill­ing ef­fect on the free­dom of the in­di­vid­ual. Whether 2018 will see the con­cerns of in­di­vid­u­als pri­ori­tised over tech­no­log­i­cal cor­po­ra­tions will en­tirely de­pend on whether hu­man be­ings choose to take it ly­ing down, or to stand up and fight.

Waymo’s fully self-driv­ing ref­er­ence ve­hi­cle, Fire­fly 1


Ap­ple’s new iPhone X of­fers a face-recog­ni­tion fea­ture YEAR IN RE­VIEW

Newspapers in English

Newspapers from UAE

© PressReader. All rights reserved.