The Lights in the Tunnel Automation Accelerating Technology and the Economy of the Future_9 potx

The economy will always create new jobs; we will never have structural unemployment as a result of ad- vancing technology This is the idea behind the “Luddite fallacy” which I cussed at

prosperity This is why capitalism has succeeded, and

oth-er economic systems have failed

In time, I think it is likely that our perceptions willshift so that we begin to truly recognize the economiccontribution that our individual marketplace decisionsmake Someday, the majority of people will be valued ineconomic terms not for what they directly produce, butfor their participation in consumption markets If we cansucceed in gradually extending that participation to thebillions of people who are now trapped in poverty—and

do so in a way that creates incentives to conserve sources and minimize environmental impact—we mayfind that the resulting consumer demand is capable of fu-eling an engine that can drive us to unprecedented eco-nomic heights

re-The Green Light

The natural cycle in the tunnel is stable and reinforcing.The vast majority of the consumers in the tunnel nowglow with a predominantly green light As time passes, thecollective intensity of the lights continues to gradually in-crease

Suddenly, we see that vast numbers of dim greenlights have begun to stream into the tunnel These newlights have barely enough intensity to make it past thethreshold, but once inside, they join the river of lights as itcourses over the panels on the tunnel walls At first, wesense that the businesses in the tunnel are straining some-what to meet this new demand, but as time passes, thecycle again strengthens The collective intensity of the light

in the tunnel begins to quickly increase We also noticethat, over time, each of the newdim lights is very graduallybecoming brighter.

As the new lights stream into the tunnel and are corporated into its natural cycle, we see that some brightwhite lights begin to pulse with new energy New panelsare now appearing in many places on the tunnel walls thatwere once dark The entrepreneurs and business owners inthe tunnel are responding to the rapidly increasing de-mand

in-As the number of lights continues to increase, thespeed with which panels are updated and the number ofnew panels appearing on the tunnel walls seems to accele-rate Although we had perceived the tunnel as being al-most infinitely vast, it now appears that the walls are com-pletely covered with panels

Even as we sense this, however, the tunnel itself gins to expand As newpanels rapidly fill the spaces on theexpanding tunnel walls, we notice that some of the brigh-test white lights are now radiating with an unprecedentedintensity Still, as the ever-increasing cycle of light contin-ues to parade through the expanding tunnel, we sensestrongly that it is the seemingly infinite number of greenlights that truly encapsulates the collective energy, enter-prise and hope of all human beings

be-A PPENDIX / F INAL

T HOUGHTS

Are the ideas presented in this book WRONG?

(Opposing arguments and my responses) … Page 210Two Questions Worth Thinking About …… Page 223Where are we now? Four Possible Cases …… Page 224The Next 10-20 years: Some Indicators to

Watch For ……… Page 227Outsmarting Marx ……… Page 237The Technology Paradox ……… Page 239Machine Intelligence and the Turing Test …… Page 241

Are the ideas presented in this book WRONG? (Opposing arguments with responses)

In this section I have listed some of the arguments thatmay be made against the ideas in this book, together with

my responses These are either conventional arguments orthings I have thought of or seen elsewhere

The economy will always create new jobs; we will never have structural unemployment as a result of ad- vancing technology

This is the idea behind the “Luddite fallacy” which I cussed at some length in Chapters 2 and 3 (see pages 95and 131) At present, I suspect that most economistswould probably be likely to agree with this statement and,therefore, disagree with what I have suggested in thisbook Here, in a nutshell, is my argument for why I think

dis-we will end up with a serious unemployment problem:

As technology advances and industries automate, thisimproves the efficiency of production and tends to makethe products and services produced by those industriesmore affordable That leaves more purchasing power inthe pockets of consumers Those consumers then go outand spend that extra money on all kinds of products andservices produced by a variety of industries Some of thoseindustries are relatively labor intensive, so they have tohire more workers to meet this demand—and so overallemployment remains stable or increases This is the reason

that, historically, technology has not led to sustained,widespread unemployment.

My argument is that accelerating automation ogy will ultimately invade many of the industries that havetraditionally been labor intensive Additionally, the process

technol-of creative destruction will destroy old industries andcreate new ones, and very few of these new industries arelikely to be labor intensive As a result, the overall econo-

my will become less labor intensive and ultimately reach a

“tipping point.” Beyond this point, the economy will nolonger be able to absorb the workers who lose jobs due toautomation: businesses will instead invest primarily inmore machines I have also argued that this process will berelentless, and if it is not addressed by some type of gov-ernment policy, we may ultimately see a precipitous drop

in consumer spending as a substantial fraction of the ulation loses confidence in its future income continuity.That, of course, would result in even more unemploymentand a downward spiral would ensue

pop-If technology resulted in unemployment, everyone would already be out of a job because technological progress has been going on for hundreds of years

This really just amounts to saying “it hasn’t happened yet,

so it will never happen.” History has proven time andagain that, where technology is concerned, something can

be impossible since the dawn of civilization and then

sud-denly, in the blink of an eye, become possible

Revolutio-nary technologies, such as the airplane and nuclear power,where all dismissed as being impossible even by preemi-nent scientists who were involved in the research that led

to their development

Today, most of us accept that technology will

contin-ue to advance and produce things that we might currently

view as impossible However, we still think too narrowly

We accept that there will be new technologies, new ucts and new industries, but most of us are not prepared

prod-to accept that all this will change the basic economic rulesthat we take for granted But why wouldn’t that be thecase? Is there a fundamental reason why accelerating tech-nology should impact nearly every aspect of our lives—but

not impact the way the economy works? As I pointed out

in Chapter 2, advancing information technology—because

it enabled the creation and distribution of financial tives—has certainly played an important role in the severi-

deriva-ty of the current economic crisis I suspect that this is just

a preview of the economic impacts that technology willhave in the future

The population is aging Once the baby boomers tire, we will have a worker shortage not unemploy- ment

re-In nearly all the developed nations (and also in China), thepopulations are rapidly aging and retirement plans are pro-jected to come under intense pressure, as too few youngworkers have to support too many older retired people.Does this imply that there is likely to be an overall short-age of workers as large numbers of older people leave theworkforce? I think it is certainly possible this may be acounteractive force that might tend to delay the impactfrom automation to a certain extent Here are some things

to consider:

§ The impact of automation on a specific job category isreally not related to the number of workers available toperform that type of job Once technology advances

to the point where a type of job can be automated, themachines to do this can easily be replicated Machines

do not need to be educated or trained, and so they arenot subject to the bottlenecks that create shortages ofworkers in fields such as nursing Therefore, in consi-dering the overall impact of machine automation, theimportant criterion is not the number of workers

available but the types of jobs that can be automated.

To the extent that there are worker shortages within aspecific job category, that would actually tend to in-crease the incentive for automation technologies to bedeveloped in that area We already see this effect in Ja-pan where significant work is being to done to develop

robots capable of assisting with care of the elderlypopulation.

§ The current consensus view seems to be that, as a sult of the 2008-9 financial crisis (and its impact on401k plans), many workers will remain in the work-force longer than originally planned This will give au-tomation technology longer to come into play beforeany worker shortages materialize

re-§ Even if the aging population does tend to retard employment, this would, of course, simply be a delay-ing factor—not a long-term solution to the problem

un-I think we can certainly expect to see worker

shortag-es in some areas, but this may very possibly be combinedwith an overall unemployment problem The danger is thatincreasing structural unemployment will unfold in parallelwith the demographic problem I suspect that most of theprojections regarding the impact of aging populations as-sume reasonably full employment among younger work-ers If this does not turn out to be the case, the situationwill obviously be much worse As I pointed out in Chapter

3, a payroll tax-based system for supporting retirementprograms might become completely unsustainable

Businesses will never fully automate because of the high initial capital investment and the lack of flexibil- ity this implies

There are some valid points here, and I think that thesefactors may, in many cases, serve to retard the process ofautomation—but in the long run they will not prevent it.Some businesses will certainly delay automation because ofthe high capital outlays required However, over time, ma-chines will become more affordable, more reliable, andmore flexible At some point, as technology advances, ma-chines will begin to outperform workers to the extent that

a non-automated business will not be competitive sider the case of online banking: it generally offers a range

Con-of services, such as automatic bill paying, that could not beoffered by a human bank teller

Keep in mind that automation offers cost benefitsbeyond simply eliminating wages There are also payrolltaxes, benefits, vacation time, management issues (and ifyou eliminate workers, you can in many cases also get rid

of the first line managers), etc There are also safety andliability issues; consider the safety advantages of a fullyautomated warehouse

The need for technical and economic flexibility mayalso tend to hold back automation for a time If a businessinvests heavily in specific machines to produce a particularproduct and then that product does not sell well, it may bestuck with equipment it does not want The obvious an-swer to that is that, in the future, automation technologywill be more flexible and easy to adapt to different prod-ucts I think the manufacturers of automation equipment

are fully aware of this issue and will build increasingly ible products.

flex-There is also the issue of economic flexibility: a ness that employs workers can lay them off in slow times,while a more automated business will be stuck with itsmachines Again, I think that, in the long run, as technolo-

busi-gy advances, businesses that don’t automate will simplynot be able to compete: and that reality will overwhelmother considerations

Another point is that both of these factors (high tal costs and the need for flexibility) may tend to push thenext wave of automation toward software applicationsgeared toward eliminating knowledge worker jobs Soft-ware is typically more flexible and has a lower up frontcost than expensive mechanical automation As I noted inChapter 2, automation of these jobs, together with off-shoring, may mean diminishing prospects for knowledgeworks and college graduates in general

capi-Machines may take over most unskilled labor, but they will never be able to do skilled or professional jobs that require lots of training and education

I think this is a dangerous misconception that stems, inpart, from a certain amount of hubris on the part ofpeople who are well educated The conventional wisdom

is that a fence has been erected within our society On thelush, garden side of the fence, are workers who havestrong educations and training These people are beneficia-ries of the information age On the toxic wasteland side ofthe fence, are relatively unskilled workers These people

have been heavily impacted by both technology and lization They often survive by stringing two or three parttime jobs together or work in lowwage jobs with fewben-efits The obvious solution is for us to find a way to offerthese people additional training—so they can hop over tothe good side of the fence.

globa-I think that the problem with this scenario is that the fence is goingto move, and it may move very rapidly The good

side of the fence is going to contract, and increasing bers of well-educated workers are going to find themselvessuddenly on the toxic side As I pointed out in Chapter 2,

num-we can expect fields such as artificial intelligence to vance rapidly in the coming years and decades Whilemany college-educated knowledge workers perform tasksthat are currently beyond the capability of computers, theynonetheless have jobs that can be broken down into a rela-tively narrow set of tasks and routines Over time, thesejobs will be increasingly subject to automation, and maywell come under significant pressure from offshoring evensooner As this trend develops, I think that the psycholog-ical impact on consumers will represent a significant dan-ger to the economy

ad-Even if I am wrong and increasing unemployment isconfined primarily to lower skill workers, the overall im-pact on our society and economy would be dramaticallynegative In the United States, two thirds of workers—andtherefore consumers—do not have a college degree Whileefforts to improve education and training are laudable, thereality is that this is not likely to offset the impact of geo-metrically accelerating technology In fact, I think that the

arguments I have put forth in this book remain relevanteven in the (I think unlikely) event that more educatedworkers are spared the impact of automation.

A great many jobs require people skills and, fore, could not be performed by machines

there-I think there are definitely some jobs that meet this dard, but probably not anywhere near enough to avoid theoverall problem that would result from automation I alsothink that people often tend to underestimate the extent towhich their job might be susceptible to automation Banktellers have “people” jobs, but that has not stopped peoplefrom using ATMs or online banking In general, consum-ers seem quite receptive to automation and self-servicetechnologies if they provide convenience

stan-Many workers whose interactions with other peopleare primarily internal to their organization might feel thatthey have jobs in which people skills are critical If youthink about it, however, that is true only if most of the

other jobs are also being done by people Once automation

takes hold within an organization, such people skills mighteventually be far less important

Finally, it is important to note that automation willhave both a direct and an indirect impact on jobs Even aperson who holds a job that is completely safe from au-tomation might still be impacted by declining consumerdemand resulting from jobs elsewhere in the economy be-ing automated

Perhaps machines will ultimately take over most of the work done by people, but this won t happen until far into the future (hundreds of years, etc.)

In general, if my argument here turns out to not be rect, then there are really two basic ways in which it can bewrong:

cor-§ The argument could be fundamentally wrong Thisimplies that the economy is capable of growing and

advancing technologically basically forever, without

creating an unemployment problem In other words,even 500 years from now, when society presumablyhas technology that is incomprehensible to us today,the economy will still provide employment for the vastmajority of people in the population

§ The argument could be premature Maybe the ideashere are basically correct, but they won’t come intoplay until far into the future However, once we ac-knowledge that at some point in the future, the econ-omy may become nearly fully automated, then as amatter of mathematics, we cannot get to that pointwithout first passing through a tipping point—beyondwhich structural unemployment will begin to be aproblem The tipping point could occur long before

we expect it

I would guess that if the trends I have presented here

do not develop, then it will most likely be because I havegotten the timing wrong However, I do think there issome pretty compelling evidence that we are fairly close to