| thE 10-Q: gEORgE giLDER

Tech­nol­ogy prophet George Gilder be­lieves Sil­i­con Val­ley’s in­no­va­tions ben­e­fit only a se­lect few.

Forbes - - CONTENTS -

The tech­nol­ogy prophet on Bell’s Law and Google. Plus: Utah’s char­i­ta­ble chem­i­cal king.

Is progress in tech­nol­ogy ac­cel­er­at­ing or de­cel­er­at­ing?

It is not ac­cel­er­at­ing. It’s con­tin­u­ing to ad­vance, of course, but I com­pletely agree with Peter Thiel that tech­nol­ogy progress is not in­evitable.

What do you mean by that?

Re­call Mar­garet Mead’s story of mariner tribes that once made their living build­ing stream­lined ca­noes to catch fish in huge vol­umes. Over time, they just for­got how to make the ca­noes. When Mead found them, they were sit­ting on the beaches look­ing at the oceans with no idea that ca­noes were the so­lu­tion to their food short­age.

But in our day, learn­ing is stored for­ever on bil­lions of de­vices. It’s not go­ing to dis­ap­pear.

We’re ac­tu­ally at risk of this kind of am­ne­sia. We for­get the real en­tre­pre­neur­ial sources of cre­ativ­ity and progress: in­ven­tion, summed up in tech­no­log­i­cal progress. It’s not good to have most of the stock mar­ket ad­vance [com­ing from] five com­pa­nies, which buy back their own stock and buy up the shares of their ri­vals. I’m talk­ing about Google, Ap­ple, Face­book, Mi­crosoft and Ama­zon.

How does big tech’s suc­cess hurt in­no­va­tion?

Their suc­cess does not rep­re­sent some fun­da­men­tal change in tech­nol­ogy. It re­flects, rather, a vast en­large­ment of gov­ern­ment reg­u­la­tions, rules that re­ally fa­vor big com­pa­nies. It re­flects their ca­pa­bil­ity of lob­by­ing and lawyer­ing and lit­i­gat­ing and find­ing a path through the mazes of rules.

Your next book is called Life Af­ter Google. Why that ti­tle?

I’m con­vinced the Google par­a­digm of mas­sive data cen­ters and ar­ti­fi­cial-in­tel­li­gence de­ter­min­ism will be tran­scended in the next era.

Re­placed by . . . ?

I’ll re­fer you to Gor­don Bell’s law: Ev­ery ten years, the rate of progress pre­dicted by Moore’s Law pro­duces a hun­dred­fold rise in com­puter cost ef­fec­tive­ness. Which then re­quires a com­pletely new com­puter ar­chi­tec­ture.

Your point be­ing that we’re now past the ten-year point of Bell’s Law and the cloud.

And lo and be­hold, a new ar­chi­tec­ture is aris­ing. It will solve the in­creas­ing con­cen­tra­tion prob­lem of the in­ter­net, which is por­ous se­cu­rity. It will be mil­lions of small data cen­ters around the world, many of them mo­bile, all us­ing cryp­tog­ra­phy and a new com­puter ar­chi­tec­ture based on blockchains and other in­ven­tions.

Why would Google not see this?

Google is trapped by its own il­lu­sion. The ad­vances in ma­chine learn­ing that Google trum­pets and preens about are re­ally just ad­vances in the speed of pro­cess­ing. When their Go-play­ing com­puter can play more Go games in a minute than the whole hu­man race has played through all his­tory, that’s not a great ad­vance in in­tel­li­gence. It’s the same in­tel­li­gence just ac­cel­er­ated to ter­a­hertz speeds. And this cre­ates this il­lu­sion for Google and oth­ers that ma­chine learn­ing can some­how gain con­scious­ness and usurp hu­mans.

Ar­ti­fi­cial in­tel­li­gence evokes both ex­cite­ment and fear. Elon Musk, for one, is fear­ful.

Musk is a tremen­dous en­tre­pre­neur and a quite stale thinker. When he starts pre­tend­ing that he’s an eth­i­cal vi­sion­ary, that hu­man life is just a sim­u­la­tion in a smarter species’ game . . .

A rather de­mor­al­iz­ing view of hu­man­ity.

It’s re­ally nuts. It’s clin­i­cally crazy. Sil­i­con Val­ley should stop try­ing to make hu­man be­ings ob­so­lete and fig­ure out how to make them more pro­duc­tive again.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.