COG­NI­TIVE COM­PUT­ING

Early adopters show that cog­ni­tive sys­tems are be­ing de­ployed now. Chan­dan Ohri says you should ex­pect to see a pro­found im­pact not only on your own or­gan­i­sa­tion but on your value chain and wider in­dus­try.

NZ Business - - CONTENTS - Chan­dan Ohri is the head of ar­ti­fi­cial in­tel­li­gence at IBM New Zealand.

Ar­ti­fi­cial in­tel­li­gence (AI), long just a prom­ise, is fi­nally here. In the last few years AI and the cog­ni­tive sys­tems that bring it to life have emerged as prac­ti­cal and pow­er­ful busi­ness plat­forms, able to un­der­stand the con­text of an in­ter­ac­tion and rea­son for it­self the best out­come. We now ex­pe­ri­ence AI in many as­pects of our daily lives, from rel­a­tively sim­ple smart­phone as­sis­tants to more com­plex ser­vices in health­care.

That AI is on the ex­ec­u­tive agenda was ev­i­dent when Con­ferenz hosted Auck­land’s first Cog­ni­tive Com­put­ing for Busi­ness Fo­rum in April, to ex­plore how cog­ni­tive sys­tems are rapidly en­ter­ing the main­stream.

The ben­e­fits of ap­ply­ing cog­ni­tive tech­nol­ogy to op­ti­mise busi­ness pro­cesses and sur­face new data in­sights is ob­vi­ous, and lead­ers are now more likely to ask how to cre­ate a true ‘cog­ni­tive busi­ness’, one that has sys­tems de­signed to en­hance dig­i­tal in­tel­li­gence ex­po­nen­tially to un­der­stand, rea­son, learn and in­ter­act.

RIGHT HERE, RIGHT NOW

So why is this blos­som­ing of cog­ni­tive com­put­ing hap­pen­ing now?

To­day, data is the most valu­able re­source of any busi­ness, en­abling the ac­tions and in­sights that em­power busi­ness dis­rup­tion.

But it can only do so if it’s fully lib­er­ated to work for you. The sheer amount and dif­fer­ent types of data be­ing cre­ated has pro­lif­er­ated and is driv­ing or­gan­i­sa­tions to seek bet­ter ways to man­age and ac­cess this data quickly.

Com­bin­ing these trends with the ca­pa­bil­ity of cog­ni­tive com­put­ing to deal with both struc­tured data (for ex­am­ple, data­bases) and un­struc­tured data, such as text and video, opens the door to a new era of ma­chines and hu­mans work­ing to­gether to come to a com­mon out­come.

Hu­man be­ings don’t cre­ate or con­sume struc­tured data, we use our own cog­ni­tive frame­works to build ma­chines, de­vices and sys­tems to gen­er­ate, col­lect and im­ple­ment ap­pro­pri­ate mod­els to un­der­stand, rea­son, learn and in­ter­act with data.

So cog­ni­tive tech­nol­ogy can “take the ro­bot out of the hu­man”, al­low­ing hu­mans to do what they are best at – mak­ing judg­ments, see­ing pat­terns, build­ing re­la­tion­ships, in­no­vat­ing and cre­at­ing.

The ex­pe­ri­ence shared by speak­ers at the re­cent fo­rum is that early wins with

cog­ni­tive tech­nolo­gies are com­ing from projects that sup­port tasks.

This was il­lus­trated by Auck­land sub­scrip­tion man­age­ment soft­ware com­pany Trans­ac­tion Ser­vices Group where the an­a­lyt­ics team has cre­ated 10,000 hours of time within its direct debit billing busi­ness by au­tomat­ing the mun­dane, repet­i­tive busi­ness tasks, free­ing peo­ple to do more in­no­va­tive work.

An­other learn­ing shared at the fo­rum, by West­pac’s ef­fi­ciency value stream lead Si­mon Page, is that suc­cess­ful cog­ni­tive projects are not tech­nol­ogy projects – they must be owned by the busi­ness and sup­ported by tech­nol­ogy.

TO­WARD TRANS­FOR­MA­TION

Al­ready there are in­di­ca­tions of the kinds of busi­ness trans­for­ma­tions cog­ni­tive com­put­ing will foster in the in­dus­trial sec­tor.

Thomas Wil­lig, head of se­cu­rity at Fletcher Build­ing, said they have be­gun in­vest­ing in fully robotic man­u­fac­tur­ing equip­ment that mon­i­tors its own us­age and sched­ules its own main­te­nance. Busi­nesses op­er­at­ing these ‘cy­ber-phys­i­cal’ pro­duc­tion sys­tems are trans­form­ing their cor­po­rate se­cu­rity func­tions as phish­ing at­tacks and other ad­vanced per­sis­tent threats are be­com­ing in­creas­ingly dif­fi­cult to de­tect.

Cog­ni­tive com­put­ing is be­ing used to sift through enor­mous amounts of data so com­pa­nies can dif­fer­en­ti­ate le­git­i­mate com­mu­ni­ca­tions from fraud­u­lent ones.

The al­ter­na­tive man­ual process re­quires peo­ple to phys­i­cally look at log files for false pos­i­tives gen­er­ated by se­cu­rity sys­tems such as in­tru­sion de­tec­tion tech­nol­ogy and fire­walls. To com­bat fraud and mit­i­gate risk it’s es­sen­tial that any se­cu­rity breaches can be iden­ti­fied fast.

In cus­tomer ser­vice, we are see­ing the evo­lu­tion of chat­bots and vir­tual as­sis­tants chang­ing and im­prov­ing the ex­pe­ri­ence be­ing de­liv­er­ing by brands and gov­ern­ment agen­cies.

IBM’s own cog­ni­tive com­put­ing plat­form, Wat­son, helps peo­ple ac­cess in­for­ma­tion and also helps ma­chines un­der­stand hu­mans. One of my favourite ex­am­ples of this in ac­tion is New Zealand founded com­pany Soul Ma­chines whose emo­tion­ally in­tel­li­gent avatars re­act in hu­man-like ways to the peo­ple they see through a cam­era.

An­other ex­am­ple from Aus­tralia’s UBank is the re­cently un­veiled ‘RoboChat’ chat­bot, also built on IBM Wat­son.

RoboChat will be avail­able on the bank’s web­site to guide po­ten­tial home buy­ers and re­fi­nancers through the home loan ap­pli­ca­tion form, pro­vid­ing real-time nat­u­ral lan­guage in­put on con­sumers’ ques­tions.

Fo­rum par­tic­i­pants were clear about their as­pi­ra­tions to use AI to be­come more agile, en­sure com­pli­ance and con­sis­tency, im­prove ser­vice and, vi­tally, in­crease cus­tomer sat­is­fac­tion.

We are in the early days of a promis­ing new tech­nol­ogy and, as with ev­ery prior world-chang­ing tech­nol­ogy, this one car­ries ma­jor im­pli­ca­tions. Many of the ques­tions it raises will re­quire time, re­search and open dis­cus­sion to an­swer.

IBM’s own prin­ci­ples for guid­ing the re­spon­si­ble de­ploy­ment of AI state that the pur­pose of these sys­tems is to aug­ment hu­man ex­per­tise, em­bed­ded in sys­tems con­trolled by hu­mans.

Sim­i­larly, be­ing trans­par­ent about the ac­cu­racy and own­er­ship of data used is es­sen­tial to en­sure wide­spread sup­port for adop­tion of AI.

An­other prin­ci­ple recog­nises that im­ple­men­ta­tion of cog­ni­tive sys­tems will change over time, as will the skills needed to take ad­van­tage of them. Our com­pany has com­mit­ted to work with in­dus­try to de­velop new learn­ing mod­els to cre­ate skills for the new kinds of work and jobs that will emerge in a cog­ni­tive econ­omy.

In New Zealand for ex­am­ple we sup­port Tech Fu­tures Lab, a new or­gan­i­sa­tion pro­vid­ing ed­u­ca­tion and train­ing for ex­ec­u­tives to de­velop skills to work with AI.

Early adopters, both Kiwi and in­ter­na­tional, show that cog­ni­tive sys­tems are be­ing de­ployed now. In some cases these are em­bed­ded into the ex­ist­ing pro­cesses of tra­di­tional or­gan­i­sa­tions, and in oth­ers, cog­ni­tive is en­abling en­tirely new busi­ness mod­els.

Ex­pect to see a pro­found im­pact not only on your own or­gan­i­sa­tion but on your value chain and wider in­dus­try. If you are ready, hold up your or­gan­i­sa­tion’s busi­ness strat­egy and chal­lenge your team to meet it us­ing AI.

“BabyX” a Soul Ma­chines dig­i­tal hu­man.

Newspapers in English

Newspapers from New Zealand

© PressReader. All rights reserved.