Will the new national strategy make the UK an AI superpower?

In the glob­al AI invest­ment, inno­va­tion and imple­men­ta­tion stakes, the UK lies in a cred­itable third place. Trail­ing the US and sec­ond-placed Chi­na, it holds a slight lead over Cana­da and South Korea, accord­ing to the Glob­al AI Index pub­lished in Decem­ber 2020 by Tor­toise Media. The moral of Aesop’s most famous fable involv­ing a tor­toise may be ‘more haste, less speed’, but West­min­ster is seek­ing to hare ahead in this race over the com­ing decade. Its nation­al AI strat­e­gy, pub­lished in Sep­tem­ber 2021, is a 10-year plan to make the coun­try an “AI super­pow­er”. But what does that mean exact­ly?

Although West­min­ster has already poured more than £2.3bn into AI ini­tia­tives since 2014, this strat­e­gy will accel­er­ate progress, promis­es Chris Philp, min­is­ter for tech­nol­o­gy and the dig­i­tal econ­o­my at the Depart­ment for Dig­i­tal, Cul­ture, Media and Sport.  

“It’s a huge­ly sig­nif­i­cant vision to help the UK strength­en its posi­tion as a glob­al sci­ence super­pow­er and seize the poten­tial of mod­ern tech­nol­o­gy to improve people’s lives and solve glob­al chal­lenges such as cli­mate change,” he declares.

The Croy­don South MP explains that the strat­e­gy has three main aims. These are to ensure that the coun­try invests in the long-term growth of AI; that the tech­nol­o­gy ben­e­fits every sec­tor of the econ­o­my and all parts of the coun­try; and that its devel­op­ment is gov­erned in a way that pro­tects the pub­lic and pre­serves the UK’s fun­da­men­tal val­ues while encour­ag­ing invest­ment and inno­va­tion. 

“We have heard repeat­ed­ly from peo­ple work­ing in and around AI that these issues are entire­ly con­nect­ed,” says Philp, hint­ing at the com­plex­i­ty of the task at hand.

What will life be like for peo­ple liv­ing and work­ing in an AI super­pow­er? “There are huge oppor­tu­ni­ties for the gov­ern­ment to cap­i­talise on this tech­nol­o­gy to improve lives,” he says. “We can deliv­er more for less and give a bet­ter expe­ri­ence as we do so. For peo­ple work­ing in the pub­lic sec­tor, it could mean a reduc­tion in the hours they spend on basic tasks, which will give them more time to find inno­v­a­tive ways of improv­ing pub­lic ser­vices.” 

Philp con­tin­ues: “For busi­ness­es, we want to ensure that there are clear rules, applied eth­i­cal prin­ci­ples and a pro-inno­va­tion reg­u­la­to­ry envi­ron­ment that can cre­ate tech pow­er­hous­es across the coun­try.”

AI will also be cru­cial in help­ing the UK to meet its legal oblig­a­tions to achieve net-zero car­bon emis­sions by 2050. Pleas­ing­ly for Philp, progress is already being made in this field. He notes that the Alan Tur­ing Insti­tute has been “explor­ing AI appli­ca­tions that could help to improve pow­er stor­age and opti­mise renew­able ener­gy deploy­ment by feed­ing solar and wind pow­er into the nation­al grid”.

The strat­e­gy has been gen­er­al­ly well received in the tech world, with most peo­ple acknowl­edg­ing that it’s an impor­tant step in the right direc­tion. But some experts have iden­ti­fied a few poten­tial short­com­ings.

Peter van der Put­ten is assis­tant pro­fes­sor of AI and cre­ative research at Lei­den Uni­ver­si­ty and direc­tor of deci­sion­ing and AI solu­tions at cloud soft­ware firm Pegasys­tems. He is “encour­aged to see a shift from broad strate­gic state­ments to more con­crete, action-ori­ent­ed rec­om­men­da­tions”, but he would have pre­ferred to see a more com­plete eth­i­cal frame­work for AI appli­ca­tion. 

“A large por­tion of the doc­u­ment focus­es on AI gov­er­nance, but it appears that a lot of the empha­sis is still on analy­sis, dis­cus­sion and pol­i­cy-mak­ing. There is less on propos­ing hard leg­is­la­tion or deter­min­ing which author­i­ty will be account­able for gov­er­nance,” van der Put­ten explains. “This is an area in which the UK will need to accel­er­ate, giv­en that both the EU and Chi­na have made rel­a­tive­ly con­crete pro­pos­als for the reg­u­la­tion of AI recent­ly.”

Liz O’Driscoll is head of inno­va­tion at Civi­ca, a sup­pli­er of soft­ware designed to improve the effi­cien­cy of pub­lic ser­vices. She believes that the UK has “made great progress so far, with many organ­i­sa­tions start­ing to embrace data stan­dards and invest in data skills. But the arti­fi­cial ele­phant in the room is human resis­tance to data-shar­ing. Pri­va­cy remains cru­cial, espe­cial­ly when it comes to cit­i­zens’ infor­ma­tion, but wider uncer­tain­ty about issues such as reg­u­la­tion, pub­lic per­cep­tion and peer endorse­ment will also prompt many in the pub­lic sec­tor to play it safe with AI.”

There are some encour­ag­ing signs that people’s gen­er­al reser­va­tions about data-shar­ing are soft­en­ing, thanks to the suc­cess of col­lab­o­ra­tive AI solu­tions dur­ing the Covid cri­sis, O’Driscoll adds. 

“Shar­ing data has been essen­tial in our defence against the virus. It has enabled key pub­lic ser­vices to stay focused on peo­ple who are most at risk,” she says. “Suc­cess sto­ries have entered the pub­lic domain, so we need to make the most of these cas­es and con­tin­ue dri­ving fur­ther pos­i­tive change.”

It’s clear that more edu­ca­tion about the ben­e­fits of data-shar­ing and work on AI ethics are required, but could a short­age of recruits prove to be the most sig­nif­i­cant chal­lenge for the nation­al AI strat­e­gy? A sur­vey pub­lished by Exper­ian in Sep­tem­ber indi­cates that more than two-thirds (68%) of UK stu­dents wrong­ly believe that they would need to earn a STEM qual­i­fi­ca­tion to stand a chance of land­ing a data-relat­ed job.

Dr Mahlet Zimeta, head of pub­lic pol­i­cy at the Open Data Insti­tute, thinks that the wide­ly held view that “the UK needs to pro­duce more peo­ple who can code” is unhelp­ful at best. 

“Although improv­ing data lit­er­a­cy is impor­tant, we’re going to need a much broad­er range of skills, includ­ing crit­i­cal think­ing,” she argues. “Lead­ers require a change of mind­set to max­imise the poten­tial of AI. At the moment, it feels as though no one wants to be the first mover, but this is why exper­i­ment­ing and being trans­par­ent about the results will dri­ve progress.”

From the government’s per­spec­tive, Philp urges both “stu­dents and busi­ness­es to equip them­selves with the skills they’ll need to take advan­tage of future devel­op­ments in AI”. For employ­ers, this will include ensur­ing that their staff “have access to suit­able train­ing and devel­op­ment oppor­tu­ni­ties”, he adds, point­ing out that the government’s online list of so-called skills boot­camps is an excel­lent place to start.Tortoise Media’s Glob­al AI Index ranks the UK fourth in the world on its sup­ply of tal­ent and third for the qual­i­ty of its research. The coun­try is a rel­a­tive lag­gard in terms of both infra­struc­ture (19th) and devel­op­ment (11th), so there is plen­ty of ground to make up on both the US and Chi­na. The nation­al AI strat­e­gy sug­gests that some haste will be required if the UK is to even keep these rivals with­in its sights. Ulti­mate­ly, though, if all goes to plan, human­i­ty stands to win.