Hu­man el­e­ment re­mains key even in the age of data

Gulf News - - COMMENT & ANALYSIS -

In 1950, the year I was born, Alan Tur­ing, who is widely con­sid­ered to be the fa­ther of com­puter science and ar­ti­fi­cial in­tel­li­gence, wrote a pa­per that opened with a ques­tion: “Can ma­chines think?” More than half a cen­tury later, the an­swer seems to be: “Yes, but ...”.

As some­one who’s been de­scribed as the ul­ti­mate don’t-con­fuse-me-with-the-facts per­son, I’m not the quintessential au­thor­ity on hu­man in­tel­li­gence, let alone ar­ti­fi­cial in­tel­li­gence. I am, how­ever, con­cerned about the num­ber of es­tab­lished busi­nesses and fledg­ling en­trepreneurs who are in­creas­ingly ea­ger to bet the farm on what the data tells them.

In other words, al­go­rithms and data should not in­ter­fere with good, old-fash­ioned, seat-of-the-pants en­trepreneur­ship, no mat­ter how dis­tant that day may be.

Here’s an ex­am­ple of what I mean: There are cer­tain in­dus­tries where big data is king. We’re start­ing to see head­lines like BlackRock Is Mak­ing Big Data Big­ger in In­sti­tu­tional In­vestor magazine, and Last Days of the Stock Picker as Money Man­agers Em­brace AI in the Fi­nan­cial Post.

Re­ports like th­ese demon­strate the reach that AI is hav­ing in busi­ness, but we must not for­get that un­like ma­chines that “think”, only hu­mans have the abil­ity to look at an idea or mar­ket op­por­tu­nity and say “to hell with the data. Maybe this doesn’t work in the­ory, but my gut tells me it will work in prac­tice.”

Such courage to make de­ci­sions that go against what the data might sug­gest are based on any com­bi­na­tion of past mis­takes, anec­do­tal mar­ket in­tel­li­gence (for which your own peo­ple in the field are a great source), or sim­ply con­trar­ian in­stincts that tell you that now is the time to throw cau­tion to the wind and ex­claim, “screw it, let’s do it” — or, rather, “screw the data, let’s do it.”

In fact, I won­der what data-driven AI would have said with re­gard to the chances of two guys named Steve — Jobs and Woz­niak — suc­cess­fully com­pet­ing with the likes of IBM in the 1970s.

And what al­go­rithm might have pre­dicted that within just 13 years of Tesla’s found­ing, Elon Musk’s up­start could sur­pass the Ford Mo­tor Com­pany in mar­ket cap­i­tal­i­sa­tion — and do it with less than 2 per cent of Ford’s unit sales?

After all, while Musk was not a “car guy”, his in­stincts told him that de­spite pre­vi­ous stum­bles in the in­dus­try, elec­tric cars would even­tu­ally be the way of the fu­ture. Also, as high-tech as his so­lar, space and au­to­mo­tive ven­tures are, it’s in­trigu­ing that in 2014 Musk warned an au­di­ence at MIT that the tech sec­tor should be “very care­ful” about AI: “With ar­ti­fi­cial in­tel­li­gence,” Musk said, “we are sum­mon­ing the de­mon”.

Air­line In­tel­li­gence

In 1984, when we started Vir­gin At­lantic Air­ways, the de­mon we con­fronted was another kind of AI — Air­line In­tel­li­gence — in the shape of deeply en­trenched gi­ants like Bri­tish Air­ways, Pan Am and TWA.

Tra­di­tional air­line met­rics were es­sen­tially scream­ing, “This is crazy! It can­not work! Don’t do it!” This is pre­cisely why we didn’t let data get in the way.

While the so-called “in­dus­try ex­perts” in­sisted that a ven­ture like ours would be ut­terly un­sus­tain­able and doom us to an early demise, we were fo­cused on think­ing dif­fer­ently about busi­ness. If we were start­ing Vir­gin At­lantic to­day, I won­der how AI would quan­tify the ben­e­fits of an in­tan­gi­ble like a flight at­ten­dant’s smile. What would the data say about win­ning cus­tomers over with courtesy and re­spect?

I’m es­pe­cially con­cerned that peo­ple who are de­vel­op­ing AI are do­ing so at the ex­pense of de­vel­op­ing hu­man tal­ent, and down­grad­ing all the in­stincts and ex­pe­ri­en­tial learn­ing that has pro­pelled us from cave dwellings to to­day’s mod­ern so­ci­eties.

When­ever pos­si­ble, com­pa­nies must em­power their peo­ple to make on-the-spot de­ci­sions and think on their feet.

If the rule books, sys­tems and data-based pro­ce­dures con­sis­tently get in the way, they should be en­cour­aged to ques­tion them rather than to blindly ac­cept the sta­tus quo.

This is a lot eas­ier for a per­son to do, but im­pos­si­ble for a ma­chine — at least for now.

In a nut­shell, ma­chines may be able to think, but they are de­void of em­pa­thy, a uniquely hu­man char­ac­ter­is­tic that goes a very long way in busi­ness and life in gen­eral. So while I am very ex­cited about the op­por­tu­ni­ties that AI will af­ford us all, in busi­ness and in daily life, I am more than happy to keep putting my faith in the bril­liance of hu­mans, too. it’s al­most

Newspapers in English

Newspapers from UAE

© PressReader. All rights reserved.