THE LIGHTS IN THE TUNNEL / 72
CopyrightedMaterial–Paperback/Kindle available @ Amazon
chess moves, then why can it not also iterate through
every known legal argument since the days when Cicero
heldforthintheRomanForum?Wouldthisbea“lesser”
formoflegalcreativity?Perhapsitwould.Butwouldthat
mattertoourlawyer’semployer?
Althoughthepracticalapplicationsofartificialintelli-
gencehavesofaremphasizedbruteforcesolutions,itisby
nomeanstruethatthisistheonlyapproachbeingtakenin
thefield.Averyimportantareaofstudyrevolvesaround
theideaofneuralnets,whichareaspecialtypeofcomputer
thatisbuiltuponamodelofthehumanbrain.Neuralnets
are currently being used in areas such as visual pattern
recognition. In the future,we canprobablyexpect some
importantadvancesinthisarea,especiallyastheengineers
whodesignneuralnetswork morecloselywithscientists
whoareuncoveringthesecretsofhowourbrainswork.
Onethingthatprobablyjumpsoutatyouaswespeak
oflawyersandradiologistsisthatthesepeoplemakealot
of money. The average radiologist in the United States
makesover$300,000.Infact,wecanreasonablysaythat
software jobs (or knowledge worker jobs) are typically
highpayingjobs.Thiscreatesaverystrongincentivefor
businessestooffshoreand,whenpossible,automatethese
jobs.Anotherpointwecanmakeisthatthereisreallyno
relationshipbetweenhowmuchtrainingisrequiredfora
humanbeing,andhowdifficultitistoautomatethejob.
Tobecomealawyeroraradiologistrequiresbothcollege
and graduate degrees,butthiswill nothold off automa-
tion. It isa relatively simple matter to program accumu-
Acceleration / 73
CopyrightedMaterial–Paperback/Kindle available @ Amazon
latedknowledgeintoanalgorithmorenteritintoadata-
base.
Forknowledgeworkers,thereisreallyadoubledose
ofbadnews.Notonlyaretheirjobspotentiallyeasierto
automate thanother job typesbecause no investment in
mechanical equipment is required; but also, the financial
incentiveforgettingridofthe jobissignificantlyhigher.
Asaresult,wecanexpectthat,inthefuture,automation
willfallheavilyonknowledgeworkersandinparticularon
highlypaidworkers.Incaseswheretechnologyisnotyet
sufficient to automate the job, offshoring is likely to be
pursuedasainterimsolution.
Given this reality, it may be that the simulation we
performedinChapter1wasactuallysomewhatconserva-
tive.Lookbackatthetablelistingtraditionaljobsonpage
59. Veryfew of thesepeopleare knowledgeworkers. In
our simulation, we assumed that automation would fall
evenlyonsomesignificantpercentageoftheaveragelights
inthetunnel.Wenowsee,however,thatautomationmay,
in fact, arrive in a relatively “top heavy” pattern.It may
wellbethatagreatmanyofthebrighterlightsinourtun-
nelwillbeamongthefirstimpacted.
Whatdoesthismeanforabusinessthatoffersprod-
ucts and services in the mass market? Clearly, it implies
thatautomationmaybepoisedtosomedayeliminatenot
justuntoldmillionsofyourpotentialcustomers—itislike-
lytohithardatyourbestcustomers.
THE LIGHTS IN THE TUNNEL / 74
CopyrightedMaterial–Paperback/Kindle available @ Amazon
Automation, Offshoring and Small Business
Wetendtothinkofautomationandoffshoringasprimari-
lyimpactingjobsinlargecorporations.Afterall,ittakesa
substantial investment to set up a relationship with an
overseasoutsourcingfirmorbringinspecializedautoma-
tionequipmentorsoftware.In thenearfuture,however,
bothofthesepracticesarelikelytobecomeincreasingac-
cessibleandinexpensiveforeventhesmallestbusinesses.
Thereisasignificanttrendtowardbreakingjobsinto
smallerpiecesorspecifictasks—whichcanthenbeeither
automatedoroffshored.Thiscapabilityisincreasinglybe-
ing offered to small businesses either as pre-packaged
softwareorthrougheasytouseonlineinterfacesoverthe
Internet.Taxpreparationisoneareawherethisapproach
is already widespread. Instead of making a large invest-
ment in sophisticated automation software, a small busi-
nessownerormanagerwillbeabletovisitawebsiteand
then rent access tothe softwareon either a per-houror
per-taskbasis.Ithinkitisverypossiblethatthesamewill
happen with task-specific offshoring. Competition be-
tweenserviceproviderswillquicklyproducelowerprices,
easiertouseonlineinterfaces,andawidervarietyofser-
vices.
Theresultwillberapidpenetrationofthesepractices
intobusinessesofallsizes.Aswesawwiththeradiologist
and the lawyer, once significant portions of jobs can be
automated, the number of workers employed will imme-
diatelybegintofall.TheU.S.SmallBusinessAdministra-
tion estimates that businesses with fewer than 500 em-
ployees have generated from 60-80 percent of all job
Acceleration / 75
CopyrightedMaterial–Paperback/Kindle available @ Amazon
growth over the past decade.
25
As it becomes easierand
cheaper for business owners to employ automation and
offshoring,wemaywellfindthatthesepracticeswillbe-
comeasignificantdragonAmerica’sprimaryjobcreation
engine.
“Hardware” Jobs and Robotics
A“hardware”jobisajobthatrequiressomeinvestmentin
mechanicalorrobotictechnologiesinorderforthejobto
be automated. The automation of hardware jobs started
long before the computer revolution. Machines used on
assemblylines, farm equipment, and heavyearthmoving
equipmentarealltechnologiesthathavedisplacedmillions
of workers in the past. As history has shown, repetitive
motion manufacturing jobs are among the easiest to au-
tomate. In fact, as I mentioned, this is how the Luddite
movementgotstartedbackin1811.However,themerger
of mechanics and computer technology into the field of
robotics will almost certainly impact an unprecedented
number and types of jobs. Whether a specific hardware
job isdifficultoreasytoautomate reallydepends onthe
combinationofskillsandmanualdexterityrequired.
Foranexampleofajobthatisverydifficulttoauto-
mate, let’s consider an auto mechanic. A mechanic ob-
viouslyrequiresagreatdealofhand-eyecoordination.He
or she has to work on thousands of different parts in a
varietyofdifferentengines,ofteninhighlyvariedstatesof
repair.Inotherwords,arobotmechanicwouldfacemany
visual recognition and manipulation problems similar to
theoneswediscussedearlierwiththerobothousekeeper.
THE LIGHTS IN THE TUNNEL / 76
CopyrightedMaterial–Paperback/Kindle available @ Amazon
In addition, the robot mechanic would require a much
higherdegreeofproblemsolvingskillthanthehousekee-
per. In fact, this diagnostic skill it not something that
couldbesolvedwithsoftwarealonebecauseitextendsto
nearlyallthehumansenses.Amechanicmaylistentothe
soundanenginemakesorevendiagnoseaproblembased
onaspecificsmell.
As things stand, we can say that becoming an auto
mechanicisprobablyaprettysafechoiceforthetimebe-
ing. But, as wesaid with the housekeeper,that doesnot
imply the job will be safe forever. Advances in robotic
technology will continue relentlessly until many of these
problems are solved. However, an even more important
factorislikelytobechangesmadetothecarsthemechan-
ic is working on. Advancing technology has already im-
pactedthewaymechanicswork;computerizeddiagnostic
toolsarenowusedtoreadfaultcodesprovidedbymicro-
processors embedded in the engine. We can expect that
thistrendwillcontinue,andthat,atsomepointinthefu-
ture,carsmaywellbedesignedspecificallytobeworked
onbyroboticmechanics.
Atruckdriverisanotherexampleofajobthatislike-
lytobeprotectedforthetimebeing,but,inthelongrun,
the reason will probably not be so much technology as
socialacceptance.Themilitaryisalreadymakingsubstan-
tialinvestmentsinautomatedtrucksthatcouldbeusedon
thebattlefield.Thesecouldbecompletelyautonomous,or
theymightbeprogrammedtosimplyfollowaleadtruck.
Similarly,manycarmanufacturerswillsoonbedeploying
collisionavoidancetechnologyincars.Thesesystemswill
Acceleration / 77
CopyrightedMaterial–Paperback/Kindle available @ Amazon
help drivers avoid mistakes that might lead to accidents;
howeverovertimetheycouldevolveintotechnologyca-
pableofdrivingthecarautonomously—justasjetairliners
nowroutinelyflyandlandwithoutassistance.
While the technology forautomated carsand trucks
may arrive, it is somewhat difficult to imagine thatmost
peoplewouldbeeagertosharetheroadwith50-tondri-
verless trucks. A second important issue would likely be
thepoweroftheTeamstersunion.Onceagain,however,I
have to givemystandard disclaimer: this doesnot mean
truckdrivingjobswillalwaysbeprotected.
The job types that are likely to be threatened fairly
soon byadvancesinroboticsare the jobs thatfit some-
wherebetweentheautomechanicandtherepetitivemo-
tion assembly line worker. As an example, consider the
shelfstockerina supermarket or chain retailstore. This
jobrequiresmoreflexibilitythanworkingonanassembly
line,butstillfallsfarshortofwhattheautomechanicfac-
es.
Thelayoutofasupermarketisstandardizedandcould
easilybeprogrammedintoacomputer.Theaislesarewide
andthefloorsaresmooth;idealterritoryforanindustrial
robot.Everyitemhasaspecificplaceontheshelves.Bar
codesmakeitasimplemattertoidentifyitems,andspecial
location markers could be placed on the shelves: a shelf
stocking robot faces few of the visual recognition issues
that challenged our housekeeping or auto mechanic ro-
bots.Designinga robot that couldmoveinventoryfrom
the stock room and place it on shelves is certainly well
withintherealmofpossibilityinthenottoodistantfuture.
THE LIGHTS IN THE TUNNEL / 78
CopyrightedMaterial–Paperback/Kindle available @ Amazon
Needless to say, if a robot can be designed to stock
shelves, then it can also be made to unload trucks and
movematerialofalltypes.
Skepticalthatrobotsmightsomedaybesteppinginto
thesejobs?Considerthatasfarbackas2005,CNETNews
Blogpublishedanarticleentitled“WhysoNervousabout
robots, Wal-Mart?”
26
Thearticlepointedout thatreports
hadsurfacedaboutWal-Marttestinginventory-takingro-
bots.Thesewouldberobotsprogrammedtonavigatethe
aislesatnightandautomaticallytakeacompletestorein-
ventory. When the CNET reporter contacted Wal-Mart
management, he received anunusually abrupt denialthat
Wal-Martwasconsideringusingrobotsinanyway.
WecantakeWal-Mart’smanagementatitswordand
assumethatit,infact,hasnoplanstouserobots.Inthe
long run, however, that won’t matter. At some point, if
oneofWal-Mart’scompetitorstriestogainanadvantage
by employing robots, then Wal-Mart and every other
competingbusinesswillreallyhavenochoicebuttofol-
lowsuit.ThepointofthisisnottovilifyWal-Martorany
otherbusinessthatmightsomedaychoosetoemployau-
tomation.Wehavetoacknowledgethat,inafreemarket
economy,everybusinesshastorespondtoitscompetitive
environment and employ the best available technologies
andprocesses.Ifitdoesnotdoso,itwillnotsurvive.
Historyhasshownthatjobautomationveryoftenin-
volves pushing a significant portion of the job onto the
customer.Automationinthecustomerserviceareaisreal-
lyself-service.ThishasbeenthecasewithATMs,automated
checkoutaislesandevenself-servegaspumps.In there-
Acceleration / 79
CopyrightedMaterial–Paperback/Kindle available @ Amazon
cently opened Future Store
27
near Düsseldorf, Germany,
in-storeretailsalesandcustomerassistanceisbeingauto-
matedviaacell-phoneinterface.Shoppersareabletoget
realtimeassistance,whileshopping,throughtheirmobile
phones.Theycanalsoscanbarcodesastheyshopand,in
thenearfuture,willbeabletopayfortheirpurchasesdi-
rectly through their phones—presumably avoiding the
checkoutislealtogether.
Thespecterofnearfullyautomatedsupermarketsand
chainretailstoresiscauseforgenuineconcern.Theseare
nowthejobsoflastresort.Thesearethejobsthatworkers
displacedfromotherindustriestakebecausethereisnoth-
ingbetteravailable.Lookbackatthetableonpage59.We
havealreadymentionedthat3.5millioncashiersarepoten-
tiallyatrisk.Thetableshowsanother4millionretailsales-
personsand2.3millionlaborersandfreight,stockandma-
terialmovers,aswellas1.7millionstockclerksandorder
fillers.Whatnewjobscouldwepossiblyfindforallthese
people?
Readanyarticleinthepopularpressaboutthefieldof
roboticsanditspotentialfutureimplications,andyouwill
almostinvariablyfindasentencepointingoutthat“inthe
future, robots will be used to perform tasks which are
dangerous for people, or jobs which people don’t really
want.” That is surely true, but it implies the somewhat
wistful assumptionthatrobots won’t beused injobs that
peopledowant.Thatisobviouslyasillyassumption.Ro-
bots,andotherformsofautomation,willbeusedinstead
THE LIGHTS IN THE TUNNEL / 80
CopyrightedMaterial–Paperback/Kindle available @ Amazon
ofpeopleassoonitbecomescosteffectiveandprofitable
forbusinessestodoso.
*
“Interface” Jobs
AthirdtypeofjobiswhatIcallan“interface”job.The
people who hold these jobs, toalarge extent, fill inthe
crackswhichcurrentlyexist betweenvariousinformation
formats and technologies. As an example, consider what
happenswhenyouapplyforahomemortgageloan.Ifyou
workwithanindependentmortgageagent,heorshewill
probablygiveyouapaperapplicationtofillout.Next,you
willneedtoretrieveandmakecopiesofyoursupporting
documentation: pay stubs, tax returns, bank statements,
insurancedocuments,etc.
Allofthis documentationwillbeonpaperoritwill
befaxedtoyou.Apropertyappraisalwillbedone,andthe
reportwillbeforwardedtotheloanagent.Oncetheloan
agentcollectseverythingtogether,heorshewillprobably
faxit all to thebank, wherealoanofficer will review it.
Ultimately,numberssuchasyoursalary,creditrating,and
the equity to loan ratio will be plugged into a computer
programandtheloanwilleitherbeapprovedordenied.
Clearly, the bulk of the labor associated with this
process is in collecting, copying, collating and faxing in-
formation.Theintellectualportionofthejob—eitherap-
provingordenyingtheloan—isprobablyalreadyessential-
lyhandledbyacomputer.Throughouttheeconomy,there
*
Formoreonroboticsanditspotentialimpactonemploymentand
on society, see Marshall Brain’s “Robotic Nation” blog at
.
Acceleration / 81
CopyrightedMaterial–Paperback/Kindle available @ Amazon
areprobablythousandsofjobsforclerksandofficework-
ersthatcontinuetoexistbecauseofthisclunkyinterface
betweenwhatexistsonpaperandwhatneedstobeina
computer.
Clearly,wecannotexpectthatthisstateofaffairswill
continueforever.Financialstatementsarealreadyavailable
online. Standard data formats are making it increasingly
easy for computers to talk directly to one another. The
“XML” standard is a very popular format that is already
widely used to move data between different businesses
overtheInternet.UsingXML,thecomputersatamanu-
facturingcompanycantalkdirectlytothecomputersbe-
longingtothecompany’ssuppliers.Thecontinuingdrive
towardpaperlessdocumentsandseamlesscommunication
islikelytoeliminatemanyofthesehumaninterfacejobsin
thecomingyears.
The Next “Killer App”
Since the beginnings of the personal computer industry,
computerhardwaresaleshaveoftenbeendrivenbyapar-
ticularsoftwareapplicationsocompellingthatithasmoti-
vatedcustomerstopurchasethemachinerequiredtorun
it.WhentheAppleIIwasintroducedin1977,it wasin-
itiallyasuccesswithinarelativelysmallgroupofcomputer
hobbyists. It wasn’t until the first electronic spreadsheet,
VisiCalc,wasdevelopedthattheAppleIIbegantogener-
ate wider interest. VisiCalc was the catalyst that helped
transformtheAppleIIfromaninterestingtoyintoatrue
businessmachine.Likewise,whentheIBMPCwasintro-
duced,Lotus1-2-3fulfilledthe“killerapp”role.Later,it
THE LIGHTS IN THE TUNNEL / 82
CopyrightedMaterial–Paperback/Kindle available @ Amazon
was graphic designand desktop publishing software that
drovetheAppleMacIntoshtosuccess.
Inrecentyears,thehighestsalesgrowthforthecom-
puterindustryhasnotbeeninhigh-enddesktopcomput-
ers butinstead in laptops and, lately, thenewernetbook
machines that provide a simple and inexpensive way to
browsetheweb.Atleastinpart,thisprobablyresultsfrom
thefactthattheaccelerationofcomputerhardwarecapa-
bilityhaslargelyoutpacedwhatisrequiredtorunmostof
thesoftwareapplicationsofinteresttotheaverageuser.If
you are primarily interested in word processing, spread-
sheetsandwebbrowsing,itmaybedifficulttojustifythe
costofa high-endcomputerwhen alowercostormore
portablemachineoffers morethanenoughpower torun
thesoftware.Likewise,itseemstobeincreasinglydifficult
for Microsoft and other software vendors to continually
addnewfeaturestodesktopproductivityapplicationsand
operating systems that are compelling enough to justify
expensiveupgrades.
YetthebusinessmodelsofbothIntelandMicrosoft
dependoncontinuingtosellevermorepowerfulproces-
sorsandneworupdatedsoftwareapplicationstotakead-
vantageofthatpower.Ifcustomersweretopermanently
turnawayfromtheideaoffasterprocessors,thebusiness
would quickly become commoditized, and Intel would
loseitscompetitiveadvantage.Forthatreason,wecanbe
surethatIntel,Microsoftandhundredsofothersoftware
companies are actively seeking the next killer app—
something that will fully leverage the vastly increased
Acceleration / 83
CopyrightedMaterial–Paperback/Kindle available @ Amazon
computerpowerthatwillbeavailableinthecomingyears
anddecades.
Ithinkthattherearegoodreasonstobelievethatthis
nextkillerapp isgoing toturnout to beartificial intelli-
gence (AI). AI applications are highlycompute intensive
andwilltakefulladvantageofallthecomputationalpower
thatnewprocessorscanoffer.NewstandaloneAIapplica-
tions will appear, but more importantly, artificial intelli-
genceislikelytobebuiltdirectlyintoexistingproductivity
applications and operating systems, as well as the enter-
prise software and database systems used by large busi-
nesses.
The market for AI software is likely to extend far
beyond the computer industry. Increasingly sophisticated
robots will demand the most advanced hardware and
software available. High-end microprocessors and AI
softwarewillalsosurelybeusedtobuildintelligenceinto
appliances,consumerdevicesandindustrialequipmentof
allkinds. Ultimately,robotsand other non-computer ap-
plicationsmaywelleclipsethepersonalcomputermarket
as the primary growth engine for leading-edge hardware
andsoftware.
Productsthatgivesomeinsightintowhatthefuture
mayholdarealreadyondisplay.Microsoftrecentlydem-
onstrateda“virtualpersonalassistant”whichappearsasa
computergeneratedpersononthescreen.Theassistantis
capable of tasks such as making airline reservations or
scheduling meetings and requires the most advanced
hardwareavailable.AccordingtotheNewYorkTimes,Mi-
crosoft’s virtual assistant can “make sophisticated deci-
THE LIGHTS IN THE TUNNEL / 84
CopyrightedMaterial–Paperback/Kindle available @ Amazon
sionsaboutthepeopleinfrontofher,judgingthingslike
theirattire,whethertheyseemimpatient,theirimportance
and theirpreferred timesfor appointments.”
28
The Times
article also quotes a Microsoft executive who speculates
thatfutureapplicationsmightincludea“medicaldoctorin
abox”thatcouldhelpwithbasicmedicalissues.
An artificial intelligence application that could dis-
pensebasic medicaladviceiscertainlya compellingidea,
especiallyin light ofthecontinuingproblemwithaccele-
rating health care costs. However, it raises an important
point.Whateducationandtrainingwouldwerequireofa
personwhodispensedsuchinformation?Wouldthisperson
needto beadoctor? Perhapsnot,butclearlythiswould
notbeoneofthelowskill, lowwagejobsthatwe often
associate with vulnerability to automation. The reality is
that there is simply little or no relationship between the
levelofeducationandtrainingrequiredforapersontodo
ajobandwhetherornotthatjobcanbeautomated.While
doctorsareprobablynotindangeroflosingtheirjobsin
theforeseeablefuture,thesamecannotbesaidformany
thousandsofknowledgeworkersandmiddlemanagersin
theprivatesector.
It’simportanttonotethat,whilehumanoidinterfaces
likeMicrosoft’svirtualassistantmakeforgreatdemonstra-
tions, the AI applications that will likely displace know-
ledgeworkerswillnotneedsuchelaborateinterfaces.They
willsimplybeworkhorseprogramsthatmaketheroutine
decisionsandperformthetasksandanalysisthatarecur-
rently the responsibility of highly paid workers sitting in
cubiclesallovertheworld.AIcapabilitymaystartoutby
Acceleration / 85
CopyrightedMaterial–Paperback/Kindle available @ Amazon
being built into the productivity applications used by
workers, but over time, it will evolve to the point that
these applications can perform much of the work auto-
nomously:AIwillbecomeatoolformanagersratherthan
workers.Theresultislikelytobesubstantialjoblossesfor
knowledge workers and a flattening of organizational
chartsthatwilleliminatelargenumbersofmiddlemanag-
ers.(Theimpactofautomationwill,ofcourse,beinaddi-
tiontothat ofoffshoring.)Manyofthesepeoplewillbe
highlyeducatedprofessionalswhohadpreviouslyassumed
thattheywere,becauseoftheirskillsandadvancededuca-
tions, beneficiaries of the trend toward an increasingly
technologicalandglobalizedworld.
*
Military Robotics
Oneofthebiggestinvestorsinroboticstechnologyisthe
Pentagon. In his recent book Wired for War: The Robotics
RevolutionandConflictinthe21stCentury,P.W.Singerpoints
out thatthe U.S. militaryexpectsrobotic technologiesto
playanincreasinglyimportantroleinconflictsof thefu-
ture.Remote-controlleddroneaircraftandbomb-defusing
groundrobotsarealreadymakingcrucialcontributionsto
thewareffortinIraqandAfghanistan.TheDefenseAd-
vanced Research Projects Agency (DARPA)—the birth-
placeoftheoriginalcomputernetworkthatledtotheIn-
ternet—now considers military robotics to be one of its
top research priorities.
29
In the coming decades, we can
anticipatefarmoreadvancedrobotsplayinganincreasing-
*
Pleasesee“MachineIntelligenceandtheTuringTest”intheAppen-
dixformoreonartificialintelligence.
THE LIGHTS IN THE TUNNEL / 86
CopyrightedMaterial–Paperback/Kindle available @ Amazon
ly autonomous role in warfare in the air, on the ground
andatsea.
Allofthismakesforaratherharshcontrastbetween
theforesightshownbythemilitaryascomparedwithcivi-
lianeconomistsandanalysts.Considertheuneventerrain
and the highly unpredictableand dynamic situationsthat
would be faced bybattlefield robots. Nowcompare that
withtheenvironmentinsideasupermarketorwarehouse.
Itseemsobviousthatdesigningrobotstoperformmuch
oftheroutineworkrequiredwithincommercialandindus-
trialjobsettingsisfarlesschallengingthanproducingau-
tonomousmilitaryrobots. TheU.S.military isrightlyin-
vestingsubstantialresourcesinstudyingthefutureimpact
ofroboticsandartificialintelligenceonthewayinwhich
future warswillbefought.Andyet,littleor no cohesive
thoughtorplanningisbeinggiventothedisruptiveimpact
thatthesetechnologieswillcertainlyhaveinthecommer-
cialsectorandontheoveralleconomy.
Robotics and Offshoring
Aswe’veshown,“software”jobsarehighlysubjecttooff-
shoringandpotentiallyalsotoautomation.Those“Hard-
ware” jobs thatrequire significant hand-eye coordination
inavariedenvironmentarecurrentlyverydifficulttofully
automate.Butwhataboutoffshoring?Canahardwarejob
beoffshored?
Infactitcan,andwearelikelytoseethisincreasingly
inthenearfuture.Asanexample,consideramanufactur-
ingassemblyline. Supposethat the highlyrepetitivejobs
have already been automated, but there remain jobs for
Acceleration / 87
CopyrightedMaterial–Paperback/Kindle available @ Amazon
skilled operators at certain key points in the production
process. How could management get rid of theseskilled
workers?
Theycouldsimplybuildaremotecontrolledrobotto
performthetask,andthenoffshorethecontrolfunction.
As we have pointed out, it is the ability to recognize a
complex visual image and then manipulate a robot arm
basedonthatimagethatisaprimarychallengepreventing
full robotic automation. Transmitting a real-time visual
imageoverseas, wherea lowpaidworkercanthenmani-
pulatethemachinery,iscertainlyalreadyfeasible.Remote
controlledrobotsarecurrentlyusedinmilitaryandpolice
applicationsthatwouldbedangerousforhumans.Wevery
likely will see suchrobotsinfactoriesand workplacesin
thenearfuture.
Nanotechnology and its Impact on Employment
One of the most exciting and high impact technological
advancesthatwecanlookforwardtointhecomingdec-
ades is in the emerging field of nanotechnology. Nanotech-
nologyisconcernedwiththemanipulationofmatteratthe
molecularoreventheatomiclevel.Inthefuture,wemay
be ableto build molecular machines:tiny inventions,far
smallerthantheheadofapin,thatcanessentiallytrans-
formmatterandcreatenearlyanythingwewantoutofthe
mostbasicingredients.
This may seem like pure fantasy until we learn that
nanotechnology is already here and has been operating
THE LIGHTS IN THE TUNNEL / 88
CopyrightedMaterial–Paperback/Kindle available @ Amazon
sincelongbeforehumanbeingswalkedtheearth.
*
Itisall
aroundusandeveninsideus.Alllivingthings,atthemost
basic level, operate under the directionof molecular ma-
chines.Weknowthatourgeneticrecipeisencodedinthe
doublehelix-shapedDNAmoleculesinthenucleusofour
cells.Buthowdoesthatrecipeget translatedinto anin-
comprehensiblycomplexorganismlikeahumanbeing?
Ifwecouldzoominandwatchtheactioninsideour
cells,we wouldseetinymolecularmachines “unzip”our
DNAmoleculesandreadportionsofourgeneticcodeina
waythatisnotunlikeacomputerscanninginabarcode.
That genetic “bar code” is then transmitted to another
areainourcells.Inatinybiologicalfactorycalledtheribo-
some,therecipecapturedinthebarcodeisagainreadby
other nano-machines that build protein molecules. It is
theseproteinmoleculesthatarethetruebuildingblocksof
life.Ourmuscletissue,thehemoglobininourred blood
cells, the insulin that we need to process sugar, the en-
zymes that digest our food—all of these and countless
thousandsofotherstructuresandchemicalsthatcomprise
ourbodies and make usfunctionareproteins.And they
areallconstructedthroughnanotechnology.
Itislikelythatthecoming“nanotech”revolutionwill
begin with the study of these existing, living machines.
Imagineateamofscientistsdescendingonanalienspace-
*
Iamreferringheretotrulyadvancednanotechnologyor“molecular
machines.”Anumberoftechniquesandprocesseswhicharecurrently
inusearereferredtoas“nanotechnology,”butthesereallyrepresent
theleadingedgeoftraditionalmaterialsscience.Itislikelytobedec-
adesbeforeadvanced nanotechnology is putto widespread,practical
use.
Acceleration / 89
CopyrightedMaterial–Paperback/Kindle available @ Amazon
craftfoundburiedintheNewMexicodesert.Theywould
beginbystudyingthisalientechnologyandattemptingto
reverse-engineer it. In time, they might begin to tinker
withthespacecraftandmakeitoperateinnewanddiffer-
entways.Eventually,theywouldunderstandthetechnolo-
gyatafundamentallevel,andtheywouldbeginusingitto
buildnewmachinesoftheirown.Thiswillquiteprobably
bethepathalongwhichnanotechnologywillevolve.
30
Nanotechnology is currentlyin itsinfancy, and itis
likelythatitwilltakedecadesbeforetrulyadvancedappli-
cations become available. Nonetheless, the field offers
enormous promise and may somedaytouch nearly every
aspectofourlives.Amazingnewtreatmentsandcuresin
thefieldofmedicine,thepossibilityofgeneratingvirtually
limitless energy from the sun, even faster and more po-
werfulcomputers,unimaginablenewpossibilitiesinmanu-
facturing—allthesethingsandmoremaycomefromna-
notechnology.
But as we contemplate these exciting possibilities,
there is another question we need to ask: willnanotech-
nologycreatejobs?Willourdisplacedcashiersandshelf-
stockers and officeclerks find employment in thenano-
technology industry? A simple application of common
senseshouldgiveustheanswer.Wespeakhereofmani-
pulatingmatteratthemolecularlevel.Thelevelofpreci-
sion requiredisobviouslybeyondanyhumanbeing, and
nanotechnologywillhavetobefullyautomated.Certainly
theremaybejobsforafewveryhighlytrainedtechnicians,
buttheideathatjobswillbecreatedforbluecollarwork-
ers is fantasy. If infact,at some pointinthe future, the
THE LIGHTS IN THE TUNNEL / 90
CopyrightedMaterial–Paperback/Kindle available @ Amazon
bulk of our traditionalmanufacturing evolves into nano-
manufacturing, the global impact on employment would
benearlybeyondmeasure.
The Future of College Education
Nearlyeveryoneagreesthatacollegedegreeisgenerallya
tickettoabrighterfuture.IntheUnitedStatesin2006,the
averageworkerwithabachelor’sdegreeearned$56,788,
31
whiletheaveragehighschoolgraduateearnedalittlemore
thanhalfthisamount,or$31,071.Workerswithgraduate
orprofessionaldegreesearnedastillhigheraveragesalary
of$82,320.Whiletheprimarymotiveforthemajorityof
individualstopursueadvancededucationisalmostcertain-
lyeconomic,wewouldallagree thateducationalsocon-
veysmanyotherbenefitsbothtotheindividualandtoso-
ciety as a whole. A person with more education seems
likelytoenjoyagenerallyricherexistence,tohaveanin-
terest in a greater variety of issues and is perhaps also
morelikelytobefocusedoncontinuingpersonalandpro-
fessional growth. A more educated society is generally a
more civil society with a lower crime rate. An educated
personislikelytohangoutinthelibrary—ratherthanon
streetcorners.
The unfortunate reality, however, is that the college
dreamislikelyatsomepointtocollidewiththetrendsin
offshoringandautomationthatwehavebeendiscussingin
this chapter.Thefactisthatcollegegraduatesveryoften
end up taking “software” jobs; they become knowledge
workers. As we have seen, these jobs—and in particular
more routine or entry level jobs—are at very high risk.
Acceleration / 91
CopyrightedMaterial–Paperback/Kindle available @ Amazon
Thedangeristhatasthesetrendsaccelerate,acollegede-
greewillbeseenincreasinglynotasatickettoaprosper-
ousfuture,butasatickettoajobthatwillverylikelyva-
porize.Atsomepointinthefuture,thehighcostofacol-
lege education, together with diminishing prospects for
collegegraduates,islikelytobegin havinganegativeim-
pactoncollegeenrollment.Thiswillbeespeciallytrueof
students coming from more modest backgrounds, but it
willhaveimpactatalllevelsofsociety.
Thisis,obviously,averyunconventionalview.Most
economists and others who study such trends would
probablystronglyargueexactlytheoppositecase:thatin
the future, a college degree will be increasingly valuable
andtherewillbestrongdemandforwell-educatedwork-
ers.
Thisisessentiallythe“skillpremium”argument—the
idea that technology is creating jobs for highly skilled
workersevenasitdestroysopportunitiesfortheunskilled.
I think the evidence clearly shows that this has indeed
been the case over the past coupleof decades, but I do
notthinkitcancontinueindefinitely.Thereasonissimple:
machines and computers are advancing in capability and
willincreasinglyinvadethe realm ofthehighlyeducated.
We’lllikelyseeevidenceofthisatsomepointintheform
ofdiminished opportunityandunemploymentamongre-
cent graduates and also among older college-educated
workerswholosejobsandareunabletofindcomparable
positions.
Wemaynotseeanactualclosingofthegapinaver-
agepay for college v. non-college graduates becauseop-
THE LIGHTS IN THE TUNNEL / 92
CopyrightedMaterial–Paperback/Kindle available @ Amazon
portunitiesforworkersofallskilllevelsarelikelytobein
decline. I am not suggesting that high school graduates
who wouldhaveotherwise gone to college will choseto
remaincompletelyunskilled,butIdothinkthereislikely
tobeamigrationtowardrelativelyskilledbluecollarjobs
ifthereisaperceptionthattheseoccupationsoffermore
security.
Asnewhighschoolgraduatesbegintoshyawayfrom
a course leading to knowledgeworkerjobs, they willin-
creasingly turn to the trades. As we have seen, jobs for
peoplelikeautomechanics,truckdrivers,plumbersandso
forthareamongthemostdifficulttoautomate.Theresult
maywellbeintensecompetitionfortheserelatively“safe”
jobs.Ashighschoolgraduateswhomightpreviouslyhave
been college-bound compete instead for trade jobs, they
will,ofcourse,endupdisplacinglessacademicallyinclined
peoplewhomayhavebeenabetterfitforthosejobs.That
willleaveevenfeweroptionsforalargenumberofwork-
ers.
We see evidence of this trend already in the daily
news.Newspapersroutinelyreportthatpeoplearespecifi-
callyseekingjobsthatcan’tbeoffshored.Muchismadeof
new“greencollarjobsthatcannotbeoutsourced.”While
this is certainly a desirable development, we have to ac-
knowledgethatthebulkofthesejobsaregoingtoinvolve
installing solar panels, wind turbines and so forth. They
aretradejobs;notjobsforcollegegraduates.
Thecosttosocietyofsuchaturnawayfromeduca-
tion would be enormous. It would damage the hopes,
dreams and expectations of our children and potentially
Acceleration / 93
CopyrightedMaterial–Paperback/Kindle available @ Amazon
rob them of things thatwe ourselves havecome to take
for granted. Those workers whose prospects were dimi-
nishedbyanewinfluxofmore“booksmart”competitors
would become even more dispirited and more likely to
turntocrimeorotherundesirablealternatives.Thisharsh
newrealitywouldfallmostheavilyonpeopleindisadvan-
tagedsectorsofthepopulation.Finally,andperhapsmost
chillingly,atrendawayfromcollegewouldrobusoftalent
wemaywellneedinthefuture.
Econometrics: Looking Backward
The majorityofmainstreameconomists, uponwhom we
dependtodetectand project trendssuchastheoneswe
have discussed here,donot seem to beparticularlycon-
cernedaboutapotentialtransitiontoanautomatedecon-
omy. Theyhold firm to the beliefthatthe economy will
continuously generate jobs within the capabilities of the
averageworker—andthatthisprocesswillcontinueinde-
finitely. I think the reason that economists cannot see
what isreallyhappeningmaybethat theyare simplytoo
buriedintheirdata.
Inrecentdecades,thefieldofeconomicshasbeenin-
creasinglytakenoverbythebranchofthefieldknownas
econometrics. Econometrics is essentially a merger of eco-
nomicsandstatistics.Econometriciansengageinthetask
ofanalyzingreamsofpasteconomicdata.Theyapplyad-
vancedstatisticaltechniquesandcreatecomplexcomputer
models.Infact,itwouldbefairtosaythateconometricsis
anotherexampleofchangethathascomeaboutasaresult
of ourvast increaseincomputationalcapability.Without
THE LIGHTS IN THE TUNNEL / 94
CopyrightedMaterial–Paperback/Kindle available @ Amazon
computers,therewouldcertainlybefarfewereconometri-
cians.
Unfortunately, the purveyors of econometrics labor
underthedelusionthattheyareeconomists,wheninfact
they are historians. Statistics is well suited to measuring
thingswhicharerelativelyconstantorwhicharechanging
gradually. It works great for baseball and for projecting
demographictrends. It doesnot, however, do well in an
environment that is that is likely to be increasingly im-
pacted bya geometric or exponential change. The graph
thatfollowsshouldhelpillustratethispoint.
Econometrics: Looking Backward
Thereasoningoftheeconometricianisillustratedby
thestraight,gradually increasing line. He orshe assumes
thatitispossibletocrunchdatafromtwo,fiveoreventen
yearsagoanddiscerntrendsthatwilllikelyholdtrueinthe
future.Theproblemiscompoundedbythefactthataba-
sic tenet of statistics is the idea that more data gives a
more reliable result. For our econometrician, more data
Projections based on past data
Technology
Time
Acceleration / 95
CopyrightedMaterial–Paperback/Kindle available @ Amazon
often meansgoing evenfurtherback into the past. So a
studybasedontenyearsworthofdatamightwellbecon-
sideredbetterthanonelookingatonlythelasttwoyears.
Theproblemwithallthisisshownbytheotherline
onthischart.This,ofcourse,isthegeometricallyincreas-
inglinethatrepresentstechnologicaladvance.Obviously,
technologyisgoingtohaveahighlydisruptiveimpacton
the assumptions of the econometricians. Why have they
not seenit?Because ageometricadvance startsout very
graduallyandthen,verysuddenly,beginstoincreasedra-
matically. In fact, by the time econometrics shows clear
evidence ofwhat is happening, itwillbe verylateinthe
game. Economistslooking at past data are alwayslooking
back at the flat (left) part of the geometrically increasing
technologyline. Prior to thepoint where the impact be-
comesobvious,thereisreallynowaypastdatacanshow
themthesteepverticalpartofthelinethatliesahead.
The Luddite Fallacy
As mentioned previously, economists have invented a
concept,namedinhonorofthe1811Ludditemovement,
called the Luddite fallacy. This line of reasoning saysthat,
while technological progress will cause some workers to
losetheir jobsas a result ofoutdated skills, any concern
thatadvancingtechnologywillleadtowidespread,increas-
ing unemployment is, in fact, a fallacy. In other words,
machineautomationwillneverleadtoeconomy-wide,sys-
temic unemployment. The reasoning offered by econo-
mists is that, asautomationincreases the productivity of
workers,itleadstolowerpricesforproductsandservices,
THE LIGHTS IN THE TUNNEL / 96
CopyrightedMaterial–Paperback/Kindle available @ Amazon
andinturn,thoselowerpricesresultinincreasedconsum-
er demand. As businesses strive to meet that increased
demand,theyramp upproduction—andthat meansnew
jobs.
Faith in the reasoning behind the Luddite fallacy is
deeplyingrainedinmostprofessionalandacademicecon-
omists.WilliamEasterlyisaprofessoratNewYorkUni-
versity and an expert in the economies of developing
countries. In his book TheElusiveQuest for Growth: Econo-
mists’ Adventures and Misadventures in the Tropics,
*
Easterly
verywellexpressestheconventionalviewthatwesetout
to challenge when we created our tunnel simulation in
Chapter1.Hismainpointisthat,asadvancingtechnology
increases productivity, prices will fall, demand will in-
crease, and the economy will therefore produce more
goodsandservices.Inotherwords,thesamenumberof
workerswillbeemployedbuttheywillproducemore.
32
The question we have to ask is: where will this in-
crease in demand come from? Who is going to step for-
*
ThisisinnowayintendedtodisparageEasterly’sbook—whichgives
manyveryusefulinsightsintotheeconomicissuesrelatedtopoverty
inthirdworldcountries.Easterly’sbookiscitedherebecauseitoffers
averycohesiveexplanationofthereasoningacceptedbythemajority
ofeconomists.Weshouldalsonotethatthebookprimarilydealswith
developingnationsinthetropics.Thesecountriesarestartingfroma
low technological level and may wellhaveexport-driveneconomies,
andsoitisnotunreasonabletoassumethattheireconomicdevelop-
ment would follow the historical trend seen in advanced countries.
However, if we extend this idea to include the future broad-based
integration of advanced technologies such as robotics and artificial
intelligenceintotheentireworldeconomy,itfallsapart.Thesetech-
nologiesaregamechangers:atsomepointinthefuture,theyaregoing
tofundamentallyaltertherelationshipbetweenworkersandmachines.