Bugged by tech­nol­ogy?

Fears over ad­vances may not be nec­es­sary

Northwest Arkansas Democrat-Gazette - - EDITORIAL PAGE - Gary Smith Gary Smith is a re­cov­er­ing jour­nal­ist liv­ing in Rogers. let­ters@nwadg.com

Here are at least some of the things I’m not sup­posed to be wor­ried about: ■ A nu­clear North Korea ■ Ex­treme weather

■ Ex­treme … peo­ple

■ Dis­eases

■ Large dogs

■ Spi­ders and/or snakes

■ The dark

And here’s what I’m sup­posed to be afraid of:

■ Ar­ti­fi­cial in­tel­li­gence.

Yep, I’m sup­posed to be OK with creepy arach­nids, but con­cerned that a com­puter pro­gram that can clue me into the lo­ca­tion of an ex­ter­mi­na­tor, al­low me to or­der bug spray in bulk and help me find a cof­fee shop (OK, not re­lated to killing spi­ders in any way, but, you know, pri­or­i­ties) is some­thing I’m sup­posed to be ter­ri­fied of.

And don’t take my word for it. No less an author­ity on crazy ideas and out­landish claims that don’t ad­here to re­al­ity than Elon Musk has said, with what we’re be­ing told is a straight face (well, he said it on Twit­ter, so, it’s kind of hard to tell), that ar­ti­fi­cial in­tel­li­gence is more dan­ger­ous than North Korea.

Ap­par­ently this was a value judg­ment and not a rank­ing. He didn’t fol­low up by say­ing AI, as it’s of­ten re­ferred to, was less dan­ger­ous than Rus­sia or Texas.

The dan­ger, I’m as­sum­ing, since 140 char­ac­ters doesn’t re­ally al­low for a lot of con­text, is that AI will start tak­ing over a lot of func­tions that are nor­mally re­served for liv­ing, breath­ing hu­man be­ings. But the worry is that what they make up for in speed and ef­fi­ciency they lack in com­pas­sion and warmth.

So, in other words, we shouldn’t trust Siri to tell us how to get to the hospi­tal, but should in­stead rely on the very nice, very em­pathic man who ex­pressed a great deal of con­cern about our well-be­ing and sent us the wrong way be­cause, while he has great per­sonal warmth, he had no sense of di­rec­tion.

We’re a na­tion of peo­ple who would trust that lat­ter over the for­mer, and who also think it’s a great idea to trust de­ter­mi­na­tions of guilt or in­no­cence to folks who couldn’t fig­ure out a way to get out of jury duty.

My chal­lenge with this is that all the hand-wring­ing is a lit­tle bit pre­ma­ture, since AI hasn’t ex­actly pro­gressed to the point where it can do a whole lot more than beat peo­ple at “Jeop­ardy” or rec­om­mend restau­rants to you.

And as in fa­vor as I am of get­ting your wor­ry­ing in early and beat­ing the Christ­mas rush (un­less what you’re wor­ried about is, in fact, the Christ­mas rush), I can’t get in too much of a lather over a sup­pos­edly malev­o­lent force whose ma­jor af­front at this point was tak­ing Euro­pean Women Po­ets for $200, Alex.

Of course, I say that now. I might be whistling a dif­fer­ent, com­puter-gen­er­ated tune com­posed of notes and sounds gleaned by ma­chines re­search­ing all hit record­ings since Bell and com­bin­ing them into a cer­tain top seller if I wind up locked in my car be­cause I’d like to get out but my cig­a­rette lighter “can’t let you do that, Dave.” And my name isn’t even Dave.

On the other hand, hav­ing my phone let me know what­ever it is I’m try­ing to do is a bad idea might ac­tu­ally be a good thing. Af­ter all, ar­ti­fi­cial or not, hav­ing some in­tel­li­gence in the area to sug­gest that any of the stupid plans I come up with might be, well, just that, is prob­a­bly a good thing. Af­ter all, the Lovely Mrs. Smith can’t be ev­ery­where.

And af­ter my daily ter­ri­fy­ing com­mute, I have to won­der if driverless cars aren’t such a bad idea. Or are at least a bit of a pref­er­ence over cars with ac­tual driv­ers, who are in re­al­ity read­ing their phones.

I mean, what­ever will I do when there isn’t a real live, breath­ing hu­man be­ing to screw up my drive-through or­der or stack the eggs on top of the bread un­der a gal­lon of milk?

Chances are we ei­ther won’t know for hun­dreds of years or will find out next Thurs­day, depend­ing on which “fu­tur­ist” (yes, you can make a liv­ing pre­dict­ing things that will hap­pen so long from now you’ll be long gone and there­fore un­touch­able ) you’d like to be­lieve.

Here’s hop­ing that, while re­cent events call into ques­tion much of our in­tel­li­gence, when AI does be­come fully in­te­grated, we’ve been smart enough to ac­cen­tu­ate the pos­i­tives and are not con­sid­er­ing “The Ter­mi­na­tor” a doc­u­men­tary.

And now if we can just do some­thing about those spi­ders …

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.