STAR-GAZERS OF THE MONTH

It’s easy to mock fu­tur­ists for play­ing guess­ing games, but both logic and data lie be­hind their de­duc­tions. Ni­cole Ko­bie ex­plores the tech­niques they use to pre­dict the fu­ture

PC Pro - - Cotents -

Ever won­dered how fu­tur­ol­o­gists plucked in­dus­try growth fig­ures seem­ingly out of the air? Ni­cole Ko­bie in­ter­views the peo­ple who do the star-gaz­ing and re­veals the tech­niques they use to pre­dict the tech fu­ture.

See­ing into the fu­ture isn’t easy. And while we love to mock those who thought the iPhone would be a dud or who crow that blockchain is the fu­ture of ev­ery­thing, there is a method to fu­tur­ists’ mad­ness. We spoke to three fu­tur­ists – al­though they don’t all call them­selves that – to find out how they peer beyond the veil of time.

Find­ing the fu­ture

Fu­tur­ists are of­ten em­ployed by com­pa­nies, tech and oth­er­wise, to help plan prod­ucts, ser­vices and or­gan­i­sa­tional struc­ture. Now a con­sul­tant, au­thor and pro­fes­sor, Brian David John­son was In­tel’s first fu­tur­ist. Due to the dif­fi­cul­ties of fabrication, the chip­maker needs to have a sense of what con­sumers want years be­fore they them­selves know. “It took them ten years to de­sign, de­velop and de­ploy a chip, so it’s of vi­tal im­por­tance for them to know to­day what peo­ple want to do ten years ahead,” he told PC Pro.

John­son’s first chip was the CE 3100, a sys­tem-on-a-chip for smart TVs. While lay­ing out the spec re­quire­ment in a 150-page doc­u­ment in 2005, he fig­ured that graph­ics would need to im­prove; rather than show­ing bars to in­di­cate an in­crease in our TV’s vol­ume, we would ex­pect bet­ter. That led In­tel to in­clude builtin graph­ics and spot a miss­ing hole in their sup­ply chain – all thanks to a fu­tur­ist who pre­dicted what TV watch­ers would ex­pect in 2015.

Tom Cheesewright is an au­thor and con­sul­tant, and de­scribes him­self as an ap­plied fu­tur­ist. That means he doesn’t spe­cialise in any par­tic­u­lar sec­tor, but has a method­ol­ogy for look­ing at any mar­ket to un­der­stand where it’s headed, cov­er­ing ar­eas from “su­per­mar­kets to su­pery­achts”, he says – though he nor­mally works for global firms and gov­ern­ments.

His work falls into three ar­eas: strate­gic, where he helps se­nior lead­er­ship and boards build longert­erm plans, of­ten in re­sponse to “a shock they have faced”; in­for­ma­tive, which means speak­ing at con­fer­ences and to the me­dia about what the fu­ture might hold; and pro­mo­tional, work­ing on mar­ket­ing and sales.

“Any pre­dic­tions I make are re­ally just a tool to help me show clients that they have to be more adapt­able, more ag­ile,” he said. “If they want to build sus­tain­able suc­cess, it prob­a­bly won’t come from do­ing more of the same thing for the next decade.”

Method(ol­ogy) in the mad­ness

Fu­tur­ists may seem to be merely guess­ing, but there is method­ol­ogy to their work. One key tech­nique is trend anal­y­sis, which looks at his­tor­i­cal data for change or growth pat­terns, and ex­trap­o­lates those lines to see what will hap­pen next. For ex­am­ple, we can pre­dict the pro­cess­ing speeds

of chips us­ing Moore’s Law, which holds that the num­ber of tran­sis­tors will dou­ble ev­ery 18 months. Com­bine that with data such as minia­tur­i­sa­tion trends or stor­age laws “and you have a sense of where that line is head­ing”, ex­plained Dr Chris Brauer, direc­tor of in­no­va­tion at the In­sti­tute of Man­age­ment Stud­ies at Gold­smiths Col­lege, Uni­ver­sity of Lon­don.

Other method­olo­gies in­clude cycli­cal pat­tern anal­y­sis (look­ing at boom and bust trends), en­vi­ron­men­tal scan­ning (in­tel­li­gence gath­er­ing on a sub­ject) and sce­nario plan­ning, which is a way of say­ing made-up sto­ries. Sce­nar­ios can in­clude “back­cast­ing”, in which you set a goal and look back at what you need to do to reach that point, or ex­per­i­men­tal de­sign, where you mock up fake prod­ucts.

One fre­quently used method is the “fu­tures cone”; plu­ral be­cause there’s not just one fu­ture dis­cussed. It’s a sim­ple idea: the en­tire cone is what is pos­si­ble – which is ev­ery­thing – and in­side are di­vi­sions be­tween what is plau­si­ble, prob­a­ble and prefer­able.

For ex­am­ple, self-driv­ing cars are pos­si­ble; that the re­quired vis­ual and com­put­ing tech­nolo­gies ex­ist are be­ing worked at makes them plau­si­ble; and they are prob­a­ble if those tech­nolo­gies are per­fected. Prefer­able de­scribes what we want to hap­pen. Will they re­place ex­ist­ing cars? Re­vamp pub­lic trans­port? Or some­thing else? That’s the prefer­able fu­ture, and this model of­fers a range of re­sults and a goal to work to­wards.

“That’s where you’re try­ing to shape the fu­ture that you want, in the de­ci­sions you make to­day,” said Brauer, not­ing that pow­er­ful tech firms can “shape the fu­ture” of their mar­kets by choos­ing to ac­quire a firm or boost in­vest­ment, or even just mar­ket tech­nolo­gies to en­cour­age them to be used in cer­tain ways.

One ben­e­fit of the fu­tures cone is flex­i­bil­ity. “You track the evo­lu­tion of it, and as you get more in­for­ma­tion or new tech­nol­ogy, you shift your ideas into dif­fer­ent col­umns,” Brauer ex­plained. A big break­through in com­put­ing vi­sion could shift self­driv­ing cars from prob­a­ble to plau­si­ble, for ex­am­ple.

There are plenty of vari­a­tions in these ideas to work with, and fu­tur­ists’ meth­ods are their sell­ing points – some even have their own pro­pri­etary tech­niques. For in­stance, Cheesewright teaches his “tools” at the Uni­ver­sity of Sal­ford and li­censes them to an­a­lysts such as KPMG. “They’re specif­i­cally de­signed to help peo­ple look at the near hori­zon and pro­duce a con­crete set of chal­lenges that can be ad­dressed, rather than a broader view of what might be in the more dis­tant fu­ture,” he said.

“My pri­mary tool helps me to work out where the big macro trends will in­ter­sect with the ex­ist­ing pres­sure points in my clients’ mar­kets. I be­lieve right now that tech­nol­ogy – in the broad­est sense – is the big­gest driver of change, so I start from there. Based on the main ef­fects that tech­nol­ogy has on ev­ery mar­ket it touches, how will it af­fect them?”

Sci-fi sto­ries

Be­ing a fu­tur­ist re­quires a head for data, but also a bright imag­i­na­tion. Brauer de­scribes one ap­proach called ex­per­i­men­tal de­sign. If you want to know how peo­ple would use a tech­nol­ogy that doesn’t yet ex­ist, fake it. Back in 2005, Brauer ran tri­als on vir­tual as­sis­tants such as Siri and Alexa – which had yet to be cre­ated.

“In the event that an AI as­sis­tant were to emerge,” he said, “what kind of char­ac­ter­is­tics would it have, and would peo­ple want it?” To an­swer that, his team sim­u­lated the ca­pa­bil­i­ties and func­tions, us­ing con­sumers in 2005 to see how con­sumers in 2018 will re­act. “Con­sumer be­hav­iour is pretty stable,” he added.

John­son uses an­other sim­i­larly imag­i­na­tive tech­nique of his own cre­ation: “sci­encefic­tion pro­to­typ­ing”. It’s es­sen­tially ef­fects-based mod­el­ling – set­ting a pre­ferred out­come and fig­ur­ing out how to reach it – but with a lit­er­ary twist: if you want to know what type of prod­uct peo­ple in the fu­ture will use, write a short story about it.

“An ef­fects-based model is a per­son in a place with a prob­lem – which is ba­si­cally how you write a story,” John­son said, not­ing that as a sci-fi au­thor, look­ing through that lens ap­peals to him. “If you’re not an ex­pert in syn­thetic bi­ol­ogy, but read sci­ence fic­tion based on fact… it gives you an on-ramp.”

Build­ing the fu­ture

Much of John­son’s work is via a tech­nique called “fu­ture­cast­ing”, es­sen­tially ef­fects-based mod­el­ling that mixes tech­ni­cal re­search with so­cial sci­ence such as cul­tural his­tory that gives com­pa­nies a range of out­comes and help reach­ing them.

“What this fu­ture­cast­ing process does is syn­the­sise dis­parate in­puts… and model the ef­fect you want to have,” he said. “Then you re­verse en­gi­neer, or back­cast, it.” In other words, he gives clients a range of out­comes, they choose the one they’d pre­fer, and then they map a path to reach that goal – it’s not merely see­ing the fu­ture, but mak­ing it.

“That’s why it’s dif­fi­cult to be wrong as a fu­tur­ist – they’re not pre­dict­ing what’s next, they’re help­ing it get built”

And that’s why it’s dif­fi­cult to be wrong as a fu­tur­ist – they’re not pre­dict­ing what’s next, they’re help­ing it get built. John­son has 50 patents to his name, sim­ply by look­ing fur­ther ahead than most. He cites famed com­puter sci­en­tist Alan Kay, say­ing: “The best way to pre­dict the fu­ture is to in­vent it.”

That said, look­ing back, each of the fu­tur­ists PC Pro spoke to have been in­cor­rect about as­pects of their pre­dic­tions – just as a flip through ten-yearold is­sues of this mag­a­zine shows we’re not al­ways right about what’s next. Brauer ad­mits he didn’t see AI de­vel­op­ing as quickly as it has, John­son says his pre­dic­tion that Big Data from tele­vi­sion would change video hasn’t come true, and Cheesewright thought we’d have week-long bat­tery life by now.

“Pre­dic­tions are just a tool for fu­tur­ists to make our clients think about what might be,” ex­plained Cheesewright. “And, if we got ev­ery­thing right, it would be a very bor­ing world in­deed.”

ABOVE In­tel fu­tur­ist, Brian David John­son ABOVE Direc­tor of in­no­va­tion, Chris Brauer ABOVE Ap­plied fu­tur­ist Tom Cheesewright

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.