Dig­i­tal Life Matthew Web­ster

The Oldie - - NEWS -

I hear ‘ar­ti­fi­cial in­tel­li­gence’ (usu­ally ‘AI’) men­tioned fre­quently th­ese days – so it’s about time I ad­dressed it here.

It’s a thorny is­sue, and I can’t de­cide if AI is, in the words of Sel­lar and Yeat­man, a good thing. Even the mighty Stephen Hawk­ing has nailed his colours firmly to the fence by declar­ing, ‘The rise of pow­er­ful AI will be ei­ther the best thing or the worst ever to hap­pen to hu­man­ity.’

So what is it? Un­for­tu­nately, AI is an in­cred­i­bly vague term but, in essence, it is the process that helps com­put­ers make de­ci­sions on our be­half, based on a set of instructions, prece­dents and rules that we give them. This might be the num­ber­plate-reader at an air­port car park or set­ting the price of an air­line ticket, or pro­vid­ing you with the re­sults of a Google search. All th­ese are man­aged by AI of one sort or an­other.

Com­put­ers don’t think as we do – yet – but they can process num­bers in­cred­i­bly quickly. AI is the frame­work we pro­vide them with to process a huge amount of data (num­bers) in a flash and come up with the an­swer that we would have done, if we had the time and en­ergy to do it.

So far, so good. But, as it be­comes more so­phis­ti­cated, AI is us­ing more and more com­plex equa­tions (th­ese are the ‘al­go­rithms’ that you also hear about) to go a step fur­ther, and try and mimic what we are do­ing when we ap­ply in­tu­ition, ex­pe­ri­ence or com­mon sense, rather than pure logic. This is much closer to ac­tual think­ing, in hu­man terms, with more scope for prob­lems.

We need to keep this in pro­por­tion. AI is in its in­fancy. Our brains, or rather our minds, are far more com­plex than the most highly de­vel­oped AI sys­tem, but only in gen­eral terms. For a highly fo­cused task – like play­ing chess – AI is now of­ten bet­ter than we are; so, make no mis­take, AI is on the move.

It is al­ready use­ful in tak­ing on bor­ing, repet­i­tive or danger­ous jobs. Mass man­u­fac­tur­ing or min­ing are good ex­am­ples but, to the alarm of some, we are also see­ing AI start­ing to do some of the jobs we thought were re­served for sen­tient hu­mans. The pro­fes­sions are in­creas­ingly ner­vous, as AI is threat­en­ing to dis­rupt them all.

Lawyers, for ex­am­ple, are es­pe­cially wor­ried that repet­i­tive and pre­dictable jobs such as prop­erty con­veyanc­ing or writ­ing wills may soon largely be un­der­taken by AI, and it won’t stop there. It’s not dif­fi­cult to imag­ing rou­tine court pro­ce­dures or con­tracts be­ing han­dled, at least in part, by robo-lawyers.

Ac­coun­tants, too, are see­ing a huge in­crease in the au­toma­tion of their pro­fes­sion, as are ar­chi­tects; there is even much talk at present of AI be­ing used by doc­tors to di­ag­nose dis­ease. It is al­ready very ev­i­dent in mass med­i­cal re­search.

One sig­nif­i­cant ben­e­fit to mankind that AI might achieve, there­fore, is to lower pro­fes­sional costs, and not be­fore time. Gen­er­ally, when labour costs rise, it leads those who pay the wages to in­crease mech­a­ni­sa­tion, and hence em­ploy fewer peo­ple. Le­gal costs in par­tic­u­lar are now so ab­surdly high that, if us clients were of­fered even a semi-au­to­matic op­tion at a lower price, I sus­pect we’d grab it with both hands.

So, is AI a good thing? While it may re­duce your le­gal fees, would you be happy if AI de­cided whether you were pro­moted or re­ceived a pay rise?

My own in­stinct is that we need some sort of con­ven­tion de­ter­min­ing the way we use AI. Some­thing like the High­way Code – cars are danger­ous but, be­cause we all tend to obey the code, we don’t usu­ally get hurt us­ing them. So it must be with AI. Wish­ful think­ing, I sus­pect.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.