AI will transform UK business — but it’s not there yet

Buoyed by pre­dic­tions from firms such as PwC that esti­mate arti­fi­cial intel­li­gence (AI) could add $15.7 tril­lion to the glob­al econ­o­my by 2030, enter­pris­es are well aware of the poten­tial of AI tech­nolo­gies to improve effi­cien­cy, accu­ra­cy and reduce over­all expen­di­ture.

Yet many are faced with sig­nif­i­cant chal­lenges in adopt­ing the tech­nol­o­gy. AI Busi­ness recent­ly sur­veyed 500 For­tune 100 C‑suite busi­ness lead­ers with the aim of deci­pher­ing what is pre­vent­ing AI from real­ly tak­ing off in busi­ness.

The very first obsta­cle iden­ti­fied by the vast major­i­ty of respon­dents may come as a sur­prise to those irri­tat­ed by the sub­stan­tial media cov­er­age sur­round­ing AI. More than 92 per cent of respon­dents iden­ti­fied the lack of under­stand­ing about AI, its capa­bil­i­ties and over­all poten­tial as the num­ber-one issue pre­vent­ing them from mak­ing the most of the tech­nol­o­gy.

A lack of understanding central to the sluggish uptake of AI

“The demys­ti­fi­ca­tion of AI is vital, oth­er­wise fear about what the tech­nol­o­gy can do to jobs, skills and peo­ple will pro­lif­er­ate,” says Sandeep Dad­lani, Mars Inc’s chief dig­i­tal offi­cer. “There is also this expec­ta­tion that AI will pro­duce mag­i­cal results in a short peri­od of time. That real­i­ty is far away – the truth is that AI is still in its infan­cy in terms of solv­ing prob­lems.”

Most organ­i­sa­tions not only lack the nec­es­sary knowl­edge, but also the right tal­ent to make a suc­cess of AI; an issue that near­ly all mem­bers of the C‑suite cit­ed as a major hur­dle. The race to recruit top AI tal­ent has reached unprece­dent­ed lev­els among the industry’s lead­ing tech ven­dors, to the point where even renowned uni­ver­si­ties are strug­gling to retain their staff on over­sub­scribed machine-learn­ing cours­es. As a result, it’s even hard­er for enter­pris­es to com­pete for AI tal­ent, which is why many organ­i­sa­tions have tak­en the route of out­sourc­ing AI to exter­nal ven­dors.

AI ownership pivotal to successful implementation

Close­ly tied to the tal­ent deficit is the piv­otal issue of own­er­ship of AI with­in the organ­i­sa­tion, iden­ti­fied by near­ly three quar­ters of all respon­dents. There’s no clear path to AI imple­men­ta­tion, with some com­pa­nies choos­ing to run small, case-by-case projects and oth­ers opt­ing for a com­pre­hen­sive AI or dig­i­tal­i­sa­tion strat­e­gy.

Research con­sul­tan­cy Trac­ti­ca rec­om­mends that cen­tralised AI con­trol is best for organ­i­sa­tions on an advanced AI path, with mul­ti­ple imple­men­ta­tions at scale, while most busi­ness­es start­ing out in AI are bet­ter off with decen­tralised projects.

For many large enter­pris­es, it falls to dig­i­tal or inno­va­tion-ori­ent­ed mem­bers of the C‑suite, such as chief infor­ma­tion offi­cers (CIOs), to evan­ge­lise for the tech­nol­o­gy with­in the wider organ­i­sa­tion, but build­ing a coali­tion of the will­ing, based on clear use-cas­es, is vital.

“It can’t just be the CIO walk­ing into the exec­u­tive meet­ing with AI,” argues Sherif Mityas, chief expe­ri­ence offi­cer at TGI Fri­days. “You need to work with those who have been through the busi­ness case. There can be lock­step between the CIO and oth­er board mem­bers who are able to see the impacts of AI in their own depart­ments and areas of respon­si­bil­i­ty.”

Data is the fuel and the starting point of AI

The short­age in AI tal­ent and lit­er­a­cy, as well as a neb­u­lous imple­men­ta­tion struc­ture, leads to anoth­er sig­nif­i­cant obsta­cle of in-house ver­sus out­sourced AI. With under­stand­ing of machine-learn­ing murky at best among exec­u­tives, select­ing the right AI solu­tion remains chal­leng­ing for busi­ness­es. Cloud ser­vice providers such as Ama­zon Web Ser­vices and Microsoft Azure offer enter­pris­es a base with which to clean their data and devel­op their own AI solu­tions, while a con­stant­ly grow­ing base of AI star­tups offers spe­cialised, use-case-ori­ent­ed appli­ca­tions.

There’s no one-size-fits-all approach, which is why many AI ven­dors rec­om­mend start­ing with a spe­cif­ic busi­ness prob­lem, rather than look­ing to apply AI for its own sake. “Busi­ness sec­tors should assess and align crit­i­cal busi­ness prob­lems with oppor­tu­ni­ties for AI tech­nol­o­gy to pro­vide val­ue through cost avoid­ance and/or new capa­bil­i­ty devel­op­ment,” says Andrea de Souza, glob­al busi­ness devel­op­ment lead for NVIDIA. Ms de Souza rec­om­mends that busi­ness­es begin with one, two or three prac­ti­cal exper­i­ments to explore if AI could have a sig­nif­i­cant impact for their organ­i­sa­tion. The key is to com­bine in-house and out­sourced approach­es to imple­men­ta­tion accord­ing to cir­cum­stances and need.

The most per­sis­tent tech­ni­cal issue that pre­vents rapid deploy­ment of AI in the enter­prise is undoubt­ed­ly organ­i­sa­tion­al data. Data is the fuel of AI and, in many cas­es, its start­ing point. With­out a good data strat­e­gy and well-organ­ised datasets, busi­ness­es will be unable to imple­ment AI effi­cient­ly or effec­tive­ly. Thank­ful­ly, enter­pris­es are grow­ing savvy to this, with many of them now invest­ing heav­i­ly in their own data sci­en­tists and experts. For those oper­at­ing on a lean busi­ness mod­el, there exist plen­ty of third-par­ty con­sul­tan­cy ser­vices to help with the heavy data lift­ing.

The UK can become a standard bearer for ethical AI

The ele­phant in the room is undoubt­ed­ly pub­lic trust. More than 80 per cent of our respon­dents remain con­cerned about cus­tomer and soci­etal respons­es towards putting AI to work. More than 60 per cent also high­light wor­ries about employ­ee atti­tudes to work­ing along­side machines. A larg­er con­ver­sa­tion around ethics, AI and the over­ar­ch­ing impact of the fourth indus­tri­al rev­o­lu­tion has yet to take place.

Thank­ful­ly, this is one area in which the UK is an emer­gent leader. Par­lia­men­tary reviews into the prospects for AI in Britain have found that, while the UK might nev­er match glob­al com­peti­tors in spend­ing, we have the infra­struc­ture and know-how nec­es­sary to become a stan­dard bear­er for eth­i­cal AI. Lord Clement-Jones, chair of the Select Com­mit­tee on Arti­fi­cial Intel­li­gence, says the UK has many of the right ingre­di­ents in place. “Busi­ness­es should focus on ensur­ing gov­ern­ment deliv­ers the cli­mate for AI in terms of get­ting the con­text right for growth,” he says. “We don’t think a spe­cial AI reg­u­la­tor is the way for­ward. It’s real­ly impor­tant that the tech indus­try gets behind all those prin­ci­ples around trans­paren­cy and explain­abil­i­ty, because we mustn’t lose pub­lic trust.”

The road to AI is long and bumpy, and unchart­ed ter­ri­to­ry. How­ev­er, the oppor­tu­ni­ty is too sig­nif­i­cant to ignore. Busi­ness­es that don’t adopt AI will be left behind in the long run. AI is already enhanc­ing human pro­duc­tiv­i­ty, yet the long jour­ney to an AI-pow­ered econ­o­my has only just begun.

A lack of understanding central to the sluggish uptake of AI

AI ownership pivotal to successful implementation

Data is the fuel and the starting point of AI