Is Skynet in­evitable?

Ex­plor­ing AI’s po­ten­tial threat and prom­ise.

The Star Malaysia - Star2 - - READS - Re­view by MArC DE FAOItE star2@ thes­tar. com. my Non­fic­tion 1. re­flec­tions of a Man by Mr Amari Soul 2. 3. 4. 5. 6. 7. 8. 9. 1. 2. 3. 4. 5. 6. 7. 8. 9. by Jo­hanna Bas­ford by Diana rikasari by Jo­hanna Bas­ford by Diana rikasari by PewDiePie by Jo­hanna Basfo

THIS book ex­plores the bur­geon­ing field of ar­ti­fi­cial in­tel­li­gence, com­monly known as AI. It is a com­pi­la­tion of sev­eral hun­dred short es­says by some of the planet’s top ex­perts on the sub­ject and quite a few non- af­fil­i­ated rather smart peo­ple who have taken the time to think about the sub­ject.

The es­says are loosely grouped, with con­cepts from one echoed, and of­ten re­futed, in an­other. While this al­lows the reader to ap­proach the sub­ject from many dif­fer­ent an­gles, there is no clear or co­he­sive vi­sion of what the ad­vent of AI will mean to hu­man­ity.

As some writ­ers point out, nar­row forms of AI are al­ready here. Google’s al­go­rithms are a good ex­am­ple – you can ask the search en­gine the day of the week on any given date in his­tory and you get the an­swer in­stantly. But is this re­ally in­tel­li­gence at work?

Re­cently, pub­lic fig­ures like Bill Gates, Stephen Hawk­ing, and Elon Musk have come out with warn­ings about the po­ten­tial threat AI poses to hu­man­ity. Ma­chines with malev­o­lent minds have been a sta­ple of sci­ence fic­tion for as long as the genre has been around ( think Ter­mi­na­tor’s Skynet, or Hal go­ing rogue in 2001: A Space Odyssey), but does the fu­ture of AI re­ally threaten hu­man­ity?

The con­clu­sion this reader draws from th­ese es­says is a con­di­tional: prob­a­bly not.

In such a short re­view, it is only pos­si­ble to barely scrape at the sur­face of the topic – the eco­nomic and mar­ket forces be­hind the drive to­wards AI would merit an es­say of its own, but for any­one who fol­lows the in­ex­orable tra­jec­tory of Moore’s un­can­nily ac­cu­rate Law, this is a book well worth read­ing, though it may lead to more ques­tions than an­swers.

( Moore’s Law arose from In­tel co- founder Gor­don Moore’s 1965 pre­dic­tion that pro­cess­ing power for com­put­ers will dou­ble ev­ery two years.)

Sim­plis­tic think­ing en­dows ma­chines with hu­man- like mo­tives – we are very quick to an­thro­po­mor­phise – but ma­chine in­telli- gence is noth­ing like hu­man in­tel­li­gence, and even re­search into “wet­ware” that repli­cates the brain’s neu­ral wiring through re­verse en­gi­neer­ing won’t pro­duce an in­tel­li­gence sim­i­lar to our own. A ma­chine that can beat a hu­man in a game of chess isn’t “in­tel­li­gent”, it’s just very fast at mak­ing cal­cu­la­tions. The ma­chine has no “de­sire” to win, it is sim­ply pro­grammed to achieve that out­come. The ma­chine doesn’t feel any ela­tion, or ex­cite­ment, or dis­ap­point­ment as the game pro­gresses.

Hu­man in­tel­li­gence is a com­plex and fickle thing. We are at the whim of our emo­tions. Given the same choice in sim­i­lar sit­u­a­tions, we will act dif­fer­ently, de­pend­ing on how we “feel” at any given mo­ment. The lev­els of dif­fer­ent neuro- trans­mit­ters in our brains will make us be­have or fol­low cour­ses of ac­tion that we might not choose in a dif­fer­ent men­tal or emo­tional state.

Our in­tel­li­gence is clouded and much of our in­ner world ex­ists on a sub- con­scious level. There are parts of our brain in­tel­li­gent enough to keep us breath­ing whether we are awake or asleep, but we can also take con­scious con­trol of our breath, or de­cide to go for a walk, or watch a movie. Much of cre­ativ­ity and in­tu­ition come from the sub­con­scious mind. Ma­chines don’t have a rich in­ner world, or thoughts, or dreams, or self- aware­ness.

In our lat­est in­car­na­tion on the path of evo­lu­tion, we have dubbed our­selves Homo sapi­ens – think­ing hu­mans – but be­fore we reached this heady state, we were Homo ha­bilis – tool- mak­ing hu­mans. The ear­li­est tools used by hu­mans were prob­a­bly ei­ther rocks or sticks. Even chim­panzees un­der­stand that a rock is more ef­fi­cient than a bare fist for crack­ing open a nut, but hu­mans have a ca­pac­ity to use things in a va­ri­ety of ways, us­ing that same rock to crack open a neigh­bour’s skull, for ex­am­ple.

Just like a rock, the tools we de­sign can be used for many dif­fer­ent things – to stream cat videos, to re­cruit sui­cide bombers, or to model novel so­lu­tions for our en­vi­ron­men­tal and eco­nomic prob­lems. The real risk of de­vel­op­ing more ca­pa­ble ma­chines is not from the ma­chines them­selves, but the use that hu­mans will make of them. But therein also lies the po­ten­tial prom­ise of a bet­ter fu­ture.

FOr week end­ing March 3, 2016: Se­cret Gar­den: An Inky trea­sure Hunt And Colour­ing Book # 88 Love Life: 88 thoughts On Love And Life – Vol 1 Enchanted For­est: An Inky Quest & Colour­ing Book # 88 Love Life: 88 thoughts On Love And Life – Vol 2 this Book Loves You Dream Cities: Colour­ing for Mind­ful­ness by rosie Good­win & Alice Chad­wick Enchanted For­est / Hu­tan Pes­ona Inky: An Inky Quest & Colour­ing Book the Amaz­ing Book Is Not On Fire: the World Of Dan And Phil by Dan how­ell & Phil Lester 10. Spark Joy: An Il­lus­trated Mas­ter Class On the Art Of Or­ga­niz­ing And tidy­ing Up

Fic­tion to Kill A Mock­ing­bird A Malaysian restau­rant In Lon­don by Tunku halim Mem­o­ries by Lang Leav Me Be­fore You by Jojo Moyes the Shock Of the Fall by nathan Filer All the Light We Can­not See the Girl With All the Gifts by M. r. Carey Wind/ Pin­ball by haruki Mu­rakami Star Wars: the Force Awak­ens Weekly list com­piled by MPh Mid Val­ley Mega­mall, Kuala Lumpur; www. mphon­line. com.

Newspapers in English

Newspapers from Malaysia

© PressReader. All rights reserved.