Raphael John-Lall

Banks are now us­ing Ar­ti­fi­cial In­tel­li­gence (AI) to de­ter­mine the cred­it­wor­thi­ness of their cus­tomers. In some coun­tries there are al­ready tri­als for self-dri­ving cars to do de­liv­ery for busi­ness­es and ro­bots have now be­come a sta­ple in pro­duc­tion lines on fac­to­ry floors.

US busi­ness mag­a­zine Forbes, in an ar­ti­cle dat­ed Ju­ly 17, 2023, spoke about how AI is “rev­o­lu­tion­is­ing” stock in­vest­ing. Ac­cord­ing to that Forbes ar­ti­cle, with AI-pow­ered tools, in­vestors can as­sess mar­ket trends, mon­i­tor com­pa­ny per­for­mance and eval­u­ate risk fac­tors more ef­fi­cient­ly than be­fore.

These are just a few ex­am­ples of how AI is im­pact­ing the world of busi­ness and fi­nance.

Man­ag­ing di­rec­tor of Ar­chi­tects of Caribbean En­ter­prise Strate­gic So­lu­tions, Ri­car­do Fras­er, told the Busi­ness Guardian that de­spite crit­i­cisms of po­ten­tial job loss­es as AI tech­nol­o­gy re­places hu­man be­ings, its pos­i­tive im­pact out­weighs its neg­a­tive ef­fects.

His com­pa­ny pro­vides con­sul­tan­cy ser­vices in dig­i­tal trans­for­ma­tion, risk man­age­ment and in­sur­ance, math­e­mat­i­cal mod­el­ling, an­a­lyt­ics and AI.

“AI presents many op­por­tu­ni­ties. Busi­ness­es and or­gan­i­sa­tions that adopt it ear­ly enough will be able to scale their pro­duc­tiv­i­ty. So, what would have tak­en a busi­ness 10 hours to do can now be done in a few min­utes. Small or­gan­i­sa­tions in the Caribbean small is­land states must and should see AI as a pos­i­tive in or­der to ex­pand their mar­kets to reach more cus­tomers glob­al­ly. To com­pete, a busi­ness own­er must be fa­mil­iar with gen­er­al AI. With AI they can do things like com­pli­ance and Gap analy­sis. There is AI that cre­ate im­ages and au­dio. There is Chat­G­PT. There are many op­tions for busi­ness­es.”

He de­fined AI as a way in which com­put­ers im­i­tate the re­al in­tel­li­gence of a hu­man mind or hu­man be­ing.

“A hu­man can make judg­ments in a sit­u­a­tion and in­di­cate what the pat­tern is. A hu­man might say there are cars on the road, the weath­er is bad and it is rain­ing, there­fore it will take longer to tra­verse from point A to point B. There­fore, that per­son will be late. Ar­ti­fi­cial In­tel­li­gence will use a col­lec­tion of math­e­mat­ics, some sta­tis­tics. It will use al­go­rithms or steps that the com­put­er will take to un­earth pat­terns of be­hav­iour and pre­dict the fu­ture. That is what Ar­ti­fi­cial In­tel­li­gence is. It en­ables the au­toma­tion of cer­tain process­es. The self-dri­ving cars are en­abled be­cause of Ar­ti­fi­cial In­tel­li­gence. On the stock mar­kets, there is trad­ing done by ro­bots and that is done via Ar­ti­fi­cial In­tel­li­gence.”

Fras­er spoke about the pos­i­tive im­pact of AI in the world of fi­nance and busi­ness.

“Banks have been us­ing Ar­ti­fi­cial In­tel­li­gence, how­ev­er it has not re­placed any jobs as yet which is good. So, banks and oth­er fi­nan­cial in­sti­tu­tions have been us­ing AI to do cred­it risk mod­el­ling which is if you ap­ply for a loan, in­stead of some­one do­ing it man­u­al­ly, the AI mod­els will pre­dict your cred­it rat­ing and will be able to grant ap­provals for that loan. This is based up­on a pre­dic­tion that loan will be re­paid. This is a much faster process than a hu­man be­ing. Cen­tral Banks around the re­gion have been slow­ly been ac­knowl­edg­ing the safe­ty and re­li­a­bil­i­ty of these pre­dic­tions made by AI mod­els.”

Ro­bot­ics Process Au­toma­tion (RPA) is an­oth­er ex­am­ple of how pop­u­lar AI is be­ing used by com­pa­nies across dif­fer­ent in­dus­tries, Fras­er said.

“A pop­u­lar ap­pli­ca­tion of RPA is where one can digi­tise a doc­u­ment, use an op­ti­cal char­ac­ter read­er to read the doc­u­ment, au­to­mat­i­cal­ly recog­nis­ing the text and even com­plete the forms. So ba­si­cal­ly with RPA com­pa­nies can com­plete forms and doc­u­ments and do da­ta en­try with­out the need for a hu­man be­ing to take those doc­u­ments to per­form those ac­tiv­i­ties. So, one can take those doc­u­ments and us­ing vi­su­al recog­ni­tion soft­ware, put it in­to some data­base, digi­tise the doc­u­ment and cre­ate data­base records. That is some­thing that ex­ists lo­cal­ly so far.”

Job loss­es

De­spite the ad­vances in pro­duc­tiv­i­ty and ef­fi­cien­cy AI will gen­er­ate in the busi­ness world, Fras­er not­ed that there is pos­si­bly a dark­er side to AI.

“Right now, some peo­ple ex­press the con­cerns that AI will re­place some jobs. I don’t think that any job will be safe from AI. I will say that that fear is well found­ed be­cause if you have ma­chine learn­ing mod­els that can do things much faster in many cas­es, AI will re­place some of the po­si­tions in those fields.

“We have the sit­u­a­tion where Wendy’s fast food restau­rant in the Unit­ed States has au­to­mat­ed the process to take cus­tomers’ or­ders and to serve cus­tomers us­ing AI.

“This means oth­ers in the fast food in­dus­try to sus­tain their com­pet­i­tive­ness may al­so have to jump on­to that tech­nol­o­gy. If there is one com­pa­ny that is gen­er­at­ing big­ger mar­gins as they are able to re­duce costs us­ing AI, there is no rea­son that oth­er com­pa­nies would not want to use this tech­nol­o­gy.”

He said an­oth­er con­cern is po­ten­tial bias in the mod­els that AI us­es.

“If you feed AI in­for­ma­tion that is of a prej­u­di­cial na­ture or in­for­ma­tion that is not cor­rect, you will come up with a mod­el that will make de­ci­sions that may vi­o­late hu­man rights. This has sparked in­ter­est and there is now a full body of re­search called ‘good AI.’

“There are spe­cial­ist po­si­tions now that check AI mod­els to see if these AI risks or mod­el­ling risks ex­ist in the AI. Imag­ine if there is a hu­man re­source man­age­ment AI that sees some­one’s job ap­pli­ca­tion and it will pre­dict who will be­come a good work­er and what their pro­duc­tiv­i­ty and per­for­mance will be like. What if this AI de­ter­mines that some peo­ple based on their eth­nic or re­li­gious back­ground should not be hired. That is a ma­jor risk of AI.”

He said it is not on­ly non-tech­ni­cal work­ers be­ing af­fect­ed by AI but the same pro­gramme de­vel­op­ers them­selves face job loss­es too as AI takes over their jobs.

“The peo­ple that build AI are al­so be­ing af­fect­ed. They are work­ing on ar­ti­fi­cial su­per in­tel­li­gence. There is a sci­ence fic­tion con­cept that there will be a point of no re­turn where the AI or ma­chines be­come aware and a lot of ex­perts be­lieve that that con­cept is al­ready be­com­ing re­al­i­ty.

“The world will reach some point where there will be AI su­per in­tel­li­gence. This AI which will be one thou­sand times more pow­er­ful than Chat­G­PT and the AI tech­nolo­gies we see to­day. There is a con­cern that this AI can con­trol the stock mar­ket.

“Even jobs that you may think will be safe from AI like teach­ing or so­cial jobs, even those jobs will be tak­en by AI.”

He con­clud­ed by say­ing that AI will have a pos­i­tive im­pact not on­ly on busi­ness­es but on so­ci­ety in gen­er­al but there must be leg­isla­tive and oth­er sys­tems that will guide how it evolves.

“There must be gov­er­nance and some sort of stan­dards for the use of AI and what is per­mit­ted. We must do AI in a sus­tain­able man­ner where we think of the wel­fare of the hu­man be­ing as well, that will be the on­ly thing to pre­vent the adop­tion of AI from get­ting out of hand to harm per­sons.”