Fuel of the future.docx
- 文档编号:7946474
- 上传时间:2023-01-27
- 格式:DOCX
- 页数:8
- 大小:92.64KB
Fuel of the future.docx
《Fuel of the future.docx》由会员分享,可在线阅读,更多相关《Fuel of the future.docx(8页珍藏版)》请在冰豆网上搜索。
Fuelofthefuture
Fuelofthefuture
Dataisgivingrisetoaneweconomy
Howisitshapingup?
ANOILrefineryisanindustrialcathedral,aplaceofpower,dramaanddarkrecesses:
ornatecrackingtowersitsgothicpinnacles,flaringgasitsstainedglass,thestenchofhydrocarbonsitsheadyincense.Datacentres,incontrast,offeralessobviousspectacle:
windowlessgreybuildingsthatboastnoheightorornament,theyseemtostretchtoinfinity.
Yetthetwohavemuchincommon.Foronething,botharestuffedwithpipes.Inrefineriesthesecollectpetrol,propaneandothercomponentsofcrudeoil,whichhavebeenseparatedbyheat.Inbigdatacentrestheytransportairtocooltensofthousandsofcomputerswhichextractvalue—patterns,predictionsandotherinsights—fromrawdigitalinformation.
Bothalsofulfilthesamerole:
producingcrucialfeedstocksfortheworldeconomy.Whethercars,plasticsormanydrugs—withoutthecomponentsofcrude,muchofmodernlifewouldnotexist.Thedistillationsofdatacentres,fortheirpart,powerallkindsofonlineservicesand,increasingly,therealworldasdevicesbecomemoreandmoreconnected.
Dataaretothiscenturywhatoilwastothelastone:
adriverofgrowthandchange.Flowsofdatahavecreatednewinfrastructure,newbusinesses,newmonopolies,newpoliticsand—crucially—neweconomics.Digitalinformationisunlikeanypreviousresource;itisextracted,refined,valued,boughtandsoldindifferentways.Itchangestherulesformarketsanditdemandsnewapproachesfromregulators.Manyabattlewillbefoughtoverwhoshouldown,andbenefitfrom,data.
Thereisanawfullottoscrapover.IDC,amarket-researchfirm,predictsthatthe“digitaluniverse”(thedatacreatedandcopiedeveryyear)willreach180zettabytes(180followedby21zeros)in2025(seechart).Pumpingitallthroughabroadbandinternetconnectionwouldtakeover450myears.Tospeedthetransferintoitsdatacentres,Amazon,ane-commercegiantwithafast-growingcloud-computingarm,usestruckspullingshippingcontainerseachpackedwithstoragedevicesholding100petabytes(amere15zeros).Toingestitall,firmsarespeedilybuildingdatarefineries.In2016Amazon,AlphabetandMicrosofttogetherrackedupnearly$32bnincapitalexpenditureandcapitalleases,upby22%fromthepreviousyear,accordingtothe WallStreetJournal.
Thequalityofdatahaschanged,too.Theyarenolongermainlystocksofdigitalinformation—databasesofnamesandotherwell-definedpersonaldata,suchasage,sexandincome.Theneweconomyismoreaboutanalysingrapidreal-timeflowsofoftenunstructureddata:
thestreamsofphotosandvideosgeneratedbyusersofsocialnetworks,thereamsofinformationproducedbycommutersontheirwaytowork,thefloodofdatafromhundredsofsensorsinajetengine.
Fromsubwaytrainsandwindturbinestotoiletseatsandtoasters—allsortsofdevicesarebecomingsourcesofdata.Theworldwillbristlewithconnectedsensors,sothatpeoplewillleaveadigitaltrailwherevertheygo,eveniftheyarenotconnectedtotheinternet.AsPaulSonderegger,abig-datastrategistatOracle,asoftware-maker,putsit:
“Datawillbetheultimateexternality:
wewillgeneratethemwhateverwedo.”
Itiswhatyouknow
Mostimportant,thevalueofdataisincreasing.FacebookandGoogleinitiallyusedthedatatheycollectedfromuserstotargetadvertisingbetter.Butinrecentyearstheyhavediscoveredthatdatacanbeturnedintoanynumberofartificial-intelligence(AI)or“cognitive”services,someofwhichwillgeneratenewsourcesofrevenue.Theseservicesincludetranslation,visualrecognitionandassessingsomeone’spersonalitybysiftingthroughtheirwritings—allofwhichcanbesoldtootherfirmstouseintheirownproducts.
Althoughsignsofthedataeconomyareeverywhere,itsshapeisonlynowbecomingclear.AnditwouldlookprettyfamiliartoJ.R.Ewing.Therearethedatamajors,agrowingnumberofwildcattersandplentyofotherfirmstryingtogetapieceoftheaction.Allareouttoexploitapowerfuleconomicenginecalledthe“data-networkeffect”—usingdatatoattractmoreusers,whothengeneratemoredata,whichhelptoimproveservices,whichattractsmoreusers.
Themajorspumpfromthemostbountifulreservoirs.Themoreuserswritecomments,“like”postsandotherwiseengagewithFacebook,forexample,themoreitlearnsaboutthoseusersandthebettertargetedtheadsonnewsfeedsbecome.Similarly,themorepeoplesearchonGoogle,thebetteritssearchresultsturnout.
Thesefirmsarealwayslookingfornewwellsofinformation.Facebookgetsitsuserstotrainsomeofitsalgorithms,forinstancewhentheyuploadandtagpicturesoffriends.Thisexplainswhyitscomputerscannowrecognisehundredsofmillionsofpeoplewith98%accuracy.Google’sdigitalbutler,called“Assistant”,getsbetteratperformingtasksandansweringquestionsthemoreitisused.
Uber,foritspart,isbestknownforitscheaptaxirides.Butifthefirmisworthanestimated$68bn,itisinpartbecauseitownsthebiggestpoolofdataaboutsupply(drivers)anddemand(passengers)forpersonaltransportation.Similarly,formostpeopleTeslaisamakeroffancyelectriccars.Butitslatestmodelscollectmountainsofdata,whichallowthefirmtooptimiseitsself-drivingalgorithmsandthenupdatethesoftwareaccordingly.Bytheendoflastyear,thefirmhadgathered1.3bnmiles-worthofdrivingdata—ordersofmagnitudemorethanWaymo,Alphabet’sself-driving-cardivision.
“Data-driven”startupsarethewildcattersoftheneweconomy:
theyprospectfordigitaloil,extractitandturnitintoclevernewservices,fromanalysingX-raysandCATscanstodeterminingwheretosprayherbicideonafield.Nexar,anIsraelistartup,hasdevisedacleverwaytousedriversasdatasources.Itsappturnstheirsmartphonesintodashcamsthattagfootageoftheirtravelsviaactionstheynormallyperform.Ifmanyunexpectedlyhitthebrakeatthesamespotontheroad,thissignalsapotholeoranotherobstacle.AscompensationforusingNexar’sapp,driversgetafreedashcamandservices,suchasadetailedreportiftheyhaveanaccident.Thefirm’sgoalistoofferallsortsofservicesthathelpdriversavoidaccidents—andforwhichthey,ortheirinsurers,willpay.Onesuchisalertsaboutpotholesorwhenacararoundablindcornersuddenlystops.
Non-techfirmsaretryingtosinkdigitalwells,too.GE,forinstance,hasdevelopedan“operatingsystemfortheindustrialinternet”,calledPredix,tohelpcustomerscontroltheirmachinery.Predixisalsoadata-collectionsystem:
itpoolsdatafromdevicesitisconnectedto,mixesthesewithotherdata,andthentrainsalgorithmsthatcanhelpimprovetheoperationsofapowerplant,whentomaintainajetenginebeforeitbreaksdownandthelike.
Asinoilmarkets,biggerdatafirmskeeptakingoversmallerones(seetable).Butanotheraspectofthedataeconomywouldlookstrangetodealersinblackgold.Oilistheworld’smosttradedcommoditybyvalue.Data,bycontrast,arehardlytradedatall,atleastnotformoney.Thatisafarcryfromwhatmanyhadinmindwhentheytalkedaboutdataasa“newassetclass”,astheWorldEconomicForum(WEF),theDavosconference-organiser-cum-think-tank,didinareportpublishedin2011.Thedataeconomy,thattermsuggests,willconsistofthrivingmarketsforbitsandbytes.Butasitstands,itismostlyacollectionofindependentsilos.
Keepittoyourself
Thisabsenceofmarketsistheresultofthesamefactorsthathavegivenrisetofirms.Allsortsof“transactioncosts”onmarkets—searchingforinformation,negotiatingdeals,enforcingcontractsandsoon—makeitsimplerandmoreefficientsimplytobringtheseactivitiesin-house.Likewise,itisoftenmoreprofitabletogenerateandusedatainsideacompanythantobuyandsellthemonanopenmarket.
Theirabundancenotwithstanding,flowsofdataarenotacommodity:
eachstreamofinformationisdifferent,intermsoftimeliness,forexample,orhowcompleteitmaybe.Thislackof“fungibility”,ineconomiclingo,makesitdifficultforbuyerstofindaspecificsetofdataandtoputapriceonit:
thevalueofeachsortishardtocomparewithotherdata.Thereisadisincentivetotradeaseachsidewillworrythatitisgettingtheshortendofthestick.
Researchershaveonlyjustbeguntodeveloppricingmethodologies,somethingGartner,aconsultancy,calls“infonomics”.Oneofitspioneers,JimShortoftheUniversityofCaliforniainSanDiego,studiescaseswhereadecisionhasbeenmadeabouthowmuchdataareworth.OnesuchinvolvesasubsidiaryofCaesarsEntertainment,agamblinggroup,thatfiledforbankruptcyin2015.Itsmostvaluableasset,at$1bn,wasdeterminedtobethedataitissaidtoholdonthe45mcustomerswhohadjoinedthecompany’scustomer-loyaltyprogrammeovertheprevious17years.
Thepricingdifficultyisanimportantreasonwhyonefirmmightfinditsimplertobuyanother,evenifitismainlyinterestedindata.Thiswasthecasein2015whenIBMreportedlyspent$2bnontheWeatherCompany,togetitshandsonmountainsofweatherdataaswellastheinfrastructuretocollectthem.Anotherfudgeisbarterdeals:
partsofBritain’sNationalHealthServiceandDeepMind,Alphabet’sAIdivision,haveagreedtoswapaccesstoanonymouspatientdataformedicalinsightsextractedfromthem.
Thefactthatdigitalinformation,unlikeoil,isalso“non-rivalrous”,meaningthatitcanbecopiedandusedbymorethanoneperson(oralgorithm)atatime,createsfurthercomplications.Itmeansthatdatacaneasilybeusedforotherpurposesthanthoseagreed.Anditaddstothec
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- Fuel of the future