CMN 148 Exam 3
What is the story about ASCI Red and the Sony PlayStation 3? *ASCI Red:* first product of US gov't's *Accelerated Strategic Computing Initiative* = was the *world's fastest supercomputer* (1996) - *calculation-intensive tasks* (simulating nuclear tests) & first to score above 1 terflop - took up 80% of a tennis court & used power of 800 homes / 800kw/hr
*Sony Playstation 3:* (2003) instead of simulating nuclear explosions = it was *devoted to drawing them / other complex graphics* in *realistic, real-time, 3D* glory (for *video-game players* not physicists) = *matched ASCI Red in performance* yet *cost $500* & used 200w by 2006 = ASCI out of service
How does Morris describe the steam revolution? *Steam Revolution = Industrial Revolution "First Machine Age"
*Steam Engine Revolution:* (James Watt invention) - allowed us to *overcome limitations of muscle power / human & animal - *generate massive amounts of useful energy / mechan. power at will* - steam engine led to --> *factories, mass production, railways, mass transp* *Steam Engine* = *most profound time of transformation world has ever seen* / *biggest & fastest transformation in entire history of world*
What is the author's point about replication of capital? *Payoff = will depend on which inputs to production = scarcest* - if digital tech create cheap subs for labor = bad time to be laborer - if digital tech can increasingly sub for capital = then capital owners shouldnt expect to earn high returns
*Threat to capital's share = from other capital* *Capital =* can be *replicated at low cost* ex. computer chips / software - *marginal value of capital = tends to fall off, even if more capital used overall* - *value of existing capital = driven down* as *new capital added cheaply at the margin* Thus: rewards earned by capitalists may not autom. grow relative to labor. Instead, shares will depend on exact details of production/distrib/governance systm.
What does today's information technology do? we might consider rising inequality less of a problem if ppl. at bottom = also seeing their lives improve thx. to tech. *Strong Bounty Argument:* focus on spread = misleading & inappropriate since bounty is more important phenom. & exits even at bottom of spread - highly skilled workers / superstars = pulling away from rest. "As long as all ppls' lives getting better, why care if some are getting a lot better?
*Today's Info. Techn* = FAVOR:* * MORE-SKILLED workers > less-skilled workers* - *increase the RETURNS to capital owners > labor* - *increase ADVANTAGES that SUPERSTARS have > everyone else* all of these trends = *increase spread* (bet. those that have jobs / dont. bet. highly skilled & edu works / less advantaged ones. bet. superstars /rest) *future techn = INCREASE SPREAD & BOOST BOUNTY*
Understand the two-by-two matrix about work discussed in the chapter. Routine Jobs = falling Non-routine Jobs = holding up
*Two-By-Two matrix:* *cognitive vs. manual & routine vs. non-routine* - *demand for work = falling most* dramatically for *routine* (regardless of whether cognit. or manual) = leads to job polarization (collapse in demand for middle-income jobs) *non-routine cognitive jobs* = (financial analysis) & *non-routine manual jobs = (hair dressing) = *held up* relatively *well*
What was Keynes' view of *technological unemployment?* What is the response that has come to dominate the discipline of economics?
*automation = could put people our of work permanently* if more gets automated *Technological Unemployment:*
What do the authors mean by "Running Out of Metric System?"
*cannot capture magnitude* of *current & future digitization* 2012 = 2.7 zettabytes digital data in world 2016 = predict Internet traffic 1.3 zettabytes = *if keep growing = will run out of metric system* (no measure that large yet) (almost out of the 'zettabyte' age...then what?)
Observe the findings about college and the VLA in detail. *FINDINGS:* - 45% of students = demonstrate NO signif. improvement on CLA after 2 yrs college - 36% did not improve AT ALL even after 4 yrs college - Previous Research = *only few decades ago, ave. college student learned great deal bet. freshman & senior years* - Today: students spend only 9% of time studying (51% on socializing/recreation/etc)
*institutions = more focused on social than academic experiences & profs rarely demand much reading/writing* (*more time spent studying* - esp. studying alone = show *great improvement on CLA*) *College Impact = largely det. by indiv. effort / involvement* in academic, interpersonal, extracurric. offerings on a campus
Note Shapiro & Varian's summary of the properties of information.
"*Info = costly to produce* but *cheap to reproduce* ex.) EU had to copy all official documents into every language rep. in EU. Once digitized = can replicate / share repeatedly
How do the authors answer Gordon's question?
(Gordon) "Is growth over?" *NO* It is just being *held back by our inability* to *process all* the *new ideas fast enough* *Combinatorial possibilities = explode* so *quickly* Soon = will be *virtually infinite number* of *potentially valuable recombinations* of *existing knowledge pieces.*
What is Varian's comment about the "sexy job," and what does the comment mean?
*"Sexy Job" in next ten yrs = *statisticians* (due to amount of data created) - *digitalization = can help us better understand the past* ex. English words ^70% bet. 1950-2000 ex. Fame = comes to ppl. more quickly / fades quicker ex. interest in Evolution = declined until Watson&Crick discovered DNA structure
Note the numbers in Jared Bernstein's research and the numbers of the economists.
*Bernstein:* (compared increases in median family income bet. 1990-2008 w changes in cost of housing, health care, college) *Family Income* = ^ 20% *Housing/College Prices* = ^ 50% *Health Care* = ^ 150%
The biggest effect on automation on workers is likely to occur where?
*Biggest effect of automation* = likely to *be on workers not in America/developed nations* BUT RATHER *in DEVELOPING NATIONS* that rely on low-cost labor for their competitive advantage - if take most of labor costs our by installing robots = competitive adv. of low wages = disappears
What is digitization?
*Digitization:**encoding info as a stream of bits* - *work of turning all types media* --> into *zeros & ones* (native computer language)
Computers are "good" at what, and "lousy" at what?
*Good* at = *following rules* *Lousy* at = *pattern recognition
Note the text's discussion of Kiva and Boston Dynamics
*Kiva:* automatons = move around warehouse safely, quickly, efficiently. - scuttle around buildings, scoot under shelving units, lift them up, bring them to humans. = acquired by Amazon (warehouse logistics) *Boston Dynamics:* robots aimed to support US troops - carry heavy loads over rough terrain "Big Dog" = looks like a mastiff & can go up steep hills, recover from ice slips, & balance heavy load while moving over uneven landscapes.
Who is losing ground in today's economy?
*MANY PEOPLE* in US/other countries = *losing ground over time* *not just relative* to others, *but in absolute terms* Income of median worker in American = lower than it was in 1999 *many PPL. FALLING BEHIND as techn. races ahead*
what do the authors describe as "two of the most important one-time events in our history?
the emergence of real, useful artificial intelligence (AI) and the connection of most of the people on the planet via a common digital network.
what is changing re: libraries, telephones, and computers
they used to be unimaginable luxuries to the developing world but not anymore
what do the authors expect regarding productivity growth in the decades ahead?...and why?
we could expect decades of complementary innovations to unfold and continue to boost productivity. However, unlike the steam engine or electricity, second machine age technologies continue to improve at a remarkably rapid exponential pace, replication their power with digital perfection and even creating more opportunities for combinatorial innovation.
what do the authors claim about environmental, social, and individual indicators of health?
with the exception of climate change virtually all environmental, social and individual indicators of health have improved over time, even as human population has increased
What are the 2 amazing events our generation will likely have the good fortune to experience?
1. *Creation of true machine intelligence* 2. *Connection of all humans via a common digital network, transforming the planet's economics* - ppl. will take advantage of this to *build techn. that astonish / delight / work for us*
How has Moore's Law kept up? Moore's Law = has *held up well over 4 decades* Today = common to use *18 months as doubling period* for general computing power
1. Law of Physics restraints = *looser in digital world* (have to do w electrons per second can be put through a channel in integrated circuit / light beams traveling thru fiber-optic cables) - at some point Moore's Law must slow down but it takes a while 2. *Brilliant Tinkering:* *finding engineering detours around roadblocks* thrown up by physics (*skirt the limitations*) ex. difficult to cram integrated circuits tighter together, so instead layered them on top of another
What are the unique properties of digital information? 1. *non-rival:* *not "used up" when* it's *used* ex.) can't use headphones he has in his ears, but the digitized music itself, you can have it too - him listening to it doesn't keep others from doing so at same time or later too
2. *close to zero marginal cost of reproduction:* extremely *cheap to* make another *copy* ex.) send file through cloud service *First Copy = expensive to create* *Additional Copies = cheap*
Most people will be what when income is distributed according to power law? *Power Law:* ex) Youtube: 20% of YouTube videos get 80% of the view - goes "viral." The other millions of videos don't get that many views.
According to *Power Law,* *MOST PEOPLE = will be BELOW AVERAGE* Over time, ave. income can increase w/o any increase in median income / without any increase in income for most people (Power Law, superstars in extremes raise the mean)
What will androids do to society and to the labor force? *Androids* = *could do absolutely EVERYTHING a human worker could do* - *productivity / output* = would *skyrocket* - *increase* in: *volume / variety / affordability of offerings* - bring *great bounty* - greatly *increase spread in income* - bring *severe dislocations* to *labor* force (*replace human workers* by staffing androids) THEREFORE *those w no assets = only have their labor to sell BUT their labor = worthless* *NO Iron Law:* that techn. progress must always be accompanied by broad job creation
Androids = can do everything human can except one skill...say *cooking* - but cuz so much competition for this 1 job, companies can offer low wages & still fill openings - superstars would still get high wages (= *androids also increase income spread*) *Better machines can sub for human workers = more human wages driven down* *Machines = complement humans when build up area where machines strong & humans weak* *Effective Production* = *require both human/machine inputs* *If unmet needs/wants in world = unemployment is warning that we arent thinking hard enough* about what needs doing. *Can do more to augment/amplify unique human capabilities* to *create new source of value instead of automating* ones that exist
How do computers' communication abilities compare to those of human beings?
Computers' communication abilities = *NOT as DEEP* as humans BUT Computers = cmn ability = *broader* ex.) Google Translate: 15 languages can be spoken into phone & can translate more than half of the 15
What is Gordon's claim about the web and e-commerce? [inset quote]
First personal computers = arrived in early 1980s w word processing, word wrap, spreadsheets Recently: *rapid development* of *web & e-commerce (after 1995) ---> process completed by (2005)*
what is the concluding point of the chapter?
GDP and productivity stats overlook much of what we value, even when using a narrow economic lens. What's more the gap between what we measure and what we value grows every time we gain access to a new good or service that never existed before, or when existing goods become free as they so often do when they are digitalized.
What is the paradox described at the beginning of the chapter?
Second Machine Age *Paradox:* *GDP = never been higher & innovation never been faster* yet *people = increasingly pessimistic about children's future living standards* *BILLIONAIRE list = more than QUADRUPLED since 2000* *Income of MEDIAN HOUSEHOLDS in America = FALLEN*
What are the most important developments in history?
The domestication of animals, the Mongol, Roman, Arabic, and Ottoman Empires. Buddha, Confucius and Socrates these men were central thinkers in the Axial Age and brought transformative schools of thought to 3 major civilizations: Indian, Chinese, and European
What kind of robots populate today's factories?
Today's factories: *highly automated* but *not full of general-purpose robots* - *dedicated / specialized machinery* - *expensive to buy, configure, reconfigure*
Who earns most of the income from the new products and services?
Top 1% = increased their earnings by 278% ('79-'07) - earned over 65% of income in US between '02-'07 Middle Class = increased earnings by just 35%
Tom Mitchell of Carnegie Mellon claims we're at the beginning of what?
We're at the beginning of: - "a *10 year period* where we're going to *transition* from *computers that can't understand language* ---> to where *computers can understand quite a bit of language"*
what is a basic income guarantee?...and what is its history?
everyone has a minimum standard of living. If people want to improve on it by working, investing, starting a company, or doing any of the other activities of the capitalist engine they certainly can, but even if they don't they will still be able to act as consumers, since they will still receive money. history: one of its early proponents was the english american political activist Thomas Paine, he advocated (1797 pamphlet agragarian justice) that everyone should be given a lump sum of money upon reaching adulthood to compensate for the unjust fact that some people were born into landowning families while others were not. MLK Jr: (1967) i am now convinced the simplest approach will prove to be the most effective--the solution to poverty is to abolish it directly by a now widely discussed measure: guaranteed income
what do the authors argue is the best way to tackle our labor force challenges?
is to grow the labor force economies
what protected ignorant or low quality sellers of products in the past?
it protected inefficient or lower-quality sellers from being unmasked by unsuspecting consumers, while geography limited competition from other sellers. No longer can the seller expect to be insulated from competitors in other locations who can deliver a better service for less.
Levy & Murnane highlighted what ares besides pattern recognition as likely to stay a human advantage?
*Complex Communication* = domain that would stay on human side ex. *conversations* critical to *effective* teaching, managing, selling, etc, that require *transfer & interpretation* of broad info, requires a human still.
What is the career advice given by Google chief economist Hal Varian? ex. data scientists / writers of mobile phone apps / genetic counselors - more ppl. having genes sequenced ex. Bill Gates went into software when saw how cheap/ubiquitous computers were becoming ex. Jeff Bezos saw opp. w low-cost online commerce & set up Amazon
"*Seek to be an indispensable complement to something that's getting cheap & plentiful* *Cognitive skills of college grads = often complements to low-cost data & cheap computer power* = helps them command premium wage
Note the Jeopardy example and point about pattern recognition and complex communication.
*Computers = combining pattern matching w complex cmn* to quite literally *beat people at their own game* ex. Jeopardy: "Watson" answered 85% of Qs accurately and beat the human contestants
Note the examples of Siri's troubles ex.) "When is the next Halley's Comet? A: You have no meetings with Halley.
*Siri:* (intelligent personal assistant) - *does not work that well* - *doesn't understand what ppl are saying sometimes* - *asks for repeat clarifications* - *gives strange / inaccurate answers* - *rude responses*
What do the authors note about changes "at the very top?"
"If you don't win silver, you lose gold." When "winner-take-all" markets become more important *income inequality will rise becuz, pay at very top = pulls away from pay in middle* - *growing gaps in wages* bet. people w / w/out college edu & between capital owners/workers = *dwarfed by bigger changes at top* Top 1% = got 2/3s of all the profits from the growth in US economy
What does Kevin Kelly observe about being paid well in the future?
"You'll be paid in the future *based on HOW WELL you WORK WITH ROBOTS"* ex. Watson robot to be programmed in medicine. While robot could make a great doctor & address large share of cases, human diagnosticians will be needed for idiosyncratic & special cases. So a *partnership between robot doctor and human doctor = more creative & robust* than if they worked alone.
What is Gordon's claim about innovation? "growth of productivity = slowed after 1970. One-time only benefits of the Great Inventions & their spin-offs = had occurred and could not happen again. After 1970 = just second-round improvements" ex. air-conditioning window units ---> central air - once main effect realized = new problem - *growth stalled & began to decline*
(Gordon) *Innovation = slowing down* "there was almost no econ. growth for 4 centuries & probably for previous millennium" prior to 1750 or when Industrial Rev. started. - once growth started = stayed upward trajectory for 200 yrs, due to 1st & 2nd Industrial Rev (electricity / internal combustion engine / indoor plumbing w running water) *greatest inventions* = *so important & far-reaching, they took 100yrs to reach their main effect*
What is Krugman's point about productivity? - so only viable way for societies to become wealthier = is for *companies/workers* to keep getting: *more output from same # inputs*
(Krugman) *Productivity is not everything...but in the long run, it is ALMOST EVERYTHING" - a *country's ability* to *improve* its *standard of living* over time = *depends* almost entirely on its *ability to raise its output per worker* (# hrs of labor it takes to produce everything)
What do the authors mean by the Second Half of the Chessboard?.....and doubling happens faster? ex. 64 chess board squares. 1 grain on one, then 2, then 4 (doubles - each square have twice as many grains as pervious) So 63 instances of doubling = 18 quintillion grains of rice - larger than Mt. Everest = *power of exponential growth* / *numbers become so big = inconceivable*
(Moore's Law) While numbers do get large in first half, we still come across them in the real world. *Second Half* though, we *lose all sense of them* *lose all sense* of *how quickly numbers appear* as *exponential growth continues* In *Second Machine Age* = *doubling happens faster & exponential growth* = much *more salient*
What is Schumpeter's point about innovation?
(Schumpeter) *Innovation = outstanding fact in economic history of capitalist society*....*largely responsible* for &most of what we would* at first sight *attribute to other factors* - fundamental *importance of innovation for growth & prosperity* *Innovation = how productivity growth happens*
Review the story about computers and chess, culminating in the Deep Blue vs. Gary Kasparov match.
*"Computer" =* *evolved from referring to job humans do --> piece of equipment* - Prediction: *machines would master human skills* / *digital computer = would be world chess champion in '68* / *machines = capable of doing any work man can do* - 1997 = *IBM computer "Deep Blue" beat world chess champ, Gary Kasparov* *Today* = *NO HUMAN can beat a mid-tier computer chess game* & now running on mobile devices/personal computers
What do the authors observe about quantity of photographs taken...and herein note the comparison of Kodak & Facebook w their economic and employment aspects? *Facebook = employs tiny fraction of the ppl. needed at Kodak* *Facebook = market value several times greater than Kodak ever did & has created at least 7 billionaires w net worth 10X greater than George Eastman's*
*3.5 TRILLION photos* have been taken since first pic in 1838 - more photos now taken every 2min than in all 19th Cent. *Facebook Users* = reached *1Billion users in 2012* (4,600 employees + 1,000 engineers) *Kodak:* (145,300 employees - 1/3 of then in Rochester, NY + thousands more via supply-chain/retail distrib. - Kodak = provided middle-class jobs for generations & created substantial share of wealth in Rochester.
There's never been a better time to be one kind of worker, and a worse time to be another...Explain.
*BETTER TIME* to be: worker with *SPECIAL SKILLS or RIGHT EDU* cuz can *USE TECHN to CREATE / CAPTURE VALUE* *WORSE TIME* to be: worker with *ORDINARY SKILLS / ABILITIES* because *computers, robots, other digital techn = acquiring these skills* at an *extraordinary rate*
How do the authors conclude the chapter? ex.) NASA did not have method to predict onset/intensity/during of a solar particle event ---> so posted its data & description of their problem on Innocentive where "non-credentialed" (ppl. w/o PhDs) can work on the problem & upload solutions no matter what their background discipline is in. - Bruce Cragin (retire radio frequency engineer) came up w the right theory even tho he had no background in astro-physics. His approach allowed NASA to predict SPE 8hrs adv. w 85% accuracy, 24hr adv w 75% accuracy
*Best way to accelerate progress = increase our capacity to test out new combinations of ideas* How: *involve more people* / power of many eyeballs (Today's digital environment = making this possible) *with more eyeballs = more digital combinations will be found* - innovations = will add up over time & be recombined *"open-innovation" / "crowdsourcing" / recombinant innovation* *ppl.* whose *expertise far away from* problem *domain = more likely to submit winning solutions*
What does "bounty" mean?
*Bounty: simultaneously more choice / greater variety /higher quality* in many areas of our lives (not simply more cheap consumer goods & empty calories) ex. more heart surgeries w/o cracking sternum ex. instant access to world's best teachers w personalized self-assessments ex. returning hearing/sight to deaf/blind
What do the authors mean by bounty and spread?
*Bounty:* *amount* of something (digital tech. ability to *replicate ideas, insights, innovations at low cost*) *Spread:* *differences* among people in *income / wealth / other important circumstances*
Note the observation of the Collegiate Learning Assessment (CLA) and college effects on the CLA.
*CLA:* test given to college students to assess their abilities in critical thinking, written cmn, problem solving, and analytic reasoning. Requires essays instead of mult. choice. - "performance task" = students get set of background documents & 90min to write essay requiring they extract info from materials given & develop POV or recommendation = *good test of ideation / pattern recognition / complex cmn*
Note the price changes of the commercial LIDAR systems described in the chapter.
*Commercial LIDAR:* (2000) *up to $35 MILLION* (2013) *$80,000* *mass production* = could allow *price to drop* to *level of a camera*
What is *DARPA*? Note the story about the car
*DARPA:* Defense Advanced Research Projects Agency founded in response to Soviet Union's launch of Sputnik satellite - 2002 Announced "Grand Challenge:" *build a completely autonomous vehicle* that could *complete 150 mile course* thru CA Mojave Desert (15 entered, few hrs in only 4 cars still operational. But by 2010, Google announced its car)
*Moore's Law* = makes computing *devices exponentially cheaper over time*
*Digitization:* makes *available*, *massive bodies of data* relevant to almost any situation *Non-Rival:* infor that can be *infinitely reproduced & reused*
What are the General Purpose Technologies (GPT), and what is the author's point about them? *GPT* = should be *pervasive / improving over time / able to spawn new innovations* - improve along *Moore's Law trajectory* - used in *every industry in world* - *lead to innovations* (autonomous cars)
*GPT* = *deep new ideas / techniques* that *have potential for important impacts on many sectors of economy* (Impacts = boosts to output due to large productivity gains) *economically significant: interrupt & accelerate normal march of econ. progress*
How is globalization transforming the economy?
*Globalization = reason why wages stagnated in US & other adv. countries *Factor Price Equalization:* *in any single market, competition* = tend to *bid prices of production* factors (labor/capital) *to a single/common price* - lower transaction in cmn costs --> one big global market
What is *ideation* and how do the authors find it relevant? How does it compare to digital labor? ex.) Scientists come up w new hypoth ex.) Journalist sniffs out a good story ex. Chef creates new dish for menu *Ideation / Creativity / Innovation* = *advantage of human over digital labor* Computers/Robots = *lousy* at *doint anything outside frame of their programming* (bad a thinking "outside the box")
*Ideation:* *coming up w NEW IDEAS / CONCEPTS* specifically: *GOOD ideas / concepts* (since computers can easily be programmed to generate new combinations of preexisting elements) *Ideation* = *where HUMANS HAVE a COMPARATIVE ADVANTAGE over machines* *Computers = for generating answers, NOT posing interesting new Qs* (that ability = still uniquely human)
What do the authors mean by "an inflection point?"
*Inflection Point:* a *bend in the curve* where *many techn*. that used to be *found only in science fiction = becoming everyday reality* ex. Star Trek: tricoders that scanned & recorded data (geological, meteorological, medical) = all on smartphones today
What is SLAM and Kinect? *SLAM: Simultaneous Localization and Mapping* = *process* of *building up* a *map of unfamiliar building as* you're *navigating thru it* - while *keeping track of where you are in it* *PROBLEM:* all approaches = *could NOT perform consistent maps for large areas* due to *increase of computational cost & uncertainties prohibited when too large* - *sensing a sizable area & immediately crunching all resulting data = PROBLEM* = prevented SLAM progress
*Kinect:* (Microsoft, sensing device for Xbox) - *keep track* of *2 active players* - *monitor: 20 joints on each* - *recognize faces / voices / gestures* - device = *made best guess about obstructed person's movements* (if players stepped in front of another) and *picked up all joints once came back into view* Accomplished this by: *digital sensors, microphones, video camera, depth perception system* Future: *Kinect = starting to solve SLAM problem* & will *extend Kinect system to implement a full SLAM approach* in future
How have other big economies - Sweden, Finland, Germany - fared (re: income inequality)?
*Main driver* of income inequality: *exponential, digital, & combinatorial change in techn.* that *undergirds our econ. system* *Sweden/Finland/Germany* = *income inequality* actually *grown more quickly* over past 20-30 yrs than in US. - Because: these countries *started w less inequality* in their *income distributions = they continue to be less unequal than US*
What is Moore's Law?
*Moore's Law:* number of transistors in an integrated circuit = *doubles approx. every year.* - complexity of min. component costs = increased at a rate of 2 per year ex. *amount of computing power* you could *buy for $1 = doubles every year*. So can buy 2X as much power/dollar in 1963 as you could in 1962
What is Moravec's paradox?
*Moravec's Paradox:* the *discovery by artificial intelligence & robotics researchers that - contrary to traditional assumptions:* *HIGH-LEVEL* reasoning = requires *very LITTLE computation* but *LOW-LEVEL sensorimotor skills* = requires *ENORMOUS* computational resources (Hard Problems = easy, Easy Problems = hard)
What do the authors say re: Levy & Murnane on pattern recognition and complex communication?
*Pattern recognition & complex cmn* = 2 broad areas where *human hold ground over digital labor* (however this may not always prove to be the case) - even in areas where digital machines have outstripeed humans, people = still have vital roles to play
Carefully read the quote by Acemoglu & Robinson. *Concern:* *econ. inequality ---> lead to greater political inequality* *Those* who are *further empowered politically = use this* to *gain greater econ. advantage / stack cards in favor / increase econ. inequality more* *Spread could actually REDUCE bounty in years to come* *Extractive properties* (service elite) = bring poverty *Origin Power/Prosperity/Poverty = democ., property rights, rule of law (inclusive ones)
*Prosperity = depends on innovation* We *waste innovation potential if* we *do not provide* a *level playing field for all* (we don't know where next Facebook/Google will come from) *If person who will make this happen* = *goes to failing school / cannot get into good university --> chances* it will become *reality = diminish* *US = generated much innovation/econ. growth last 200yrs cuz it rewarded innovation/investment* *Supportive political arrangements* (inclusive political institutions) = *prevented elite* / narrow groups from *monopolizing pol. power* & using it for own *benefit/expense of society*
What is the second machine age? Mental Power = as important for progress&development = as Physical Power Steam Engine = relieved physical/muscle power 2nd Machine Age (computer) = relieved mental power
*Second Machine Age:* *Computers & other digital advances* = doing for *mental power what steam engine did for muscle power* - *computers & digital technology* - *Mental Power =* important for *progress & development* (mastering our physical & intellectual environ. to get things done) - boost to mental power = boost to humanity
In a superstar economy, what is the shape of the distribution of income? (in contrast w a "normal curve" where few ppl. are at the extremes) If normal distribution (few ppl. in extremes) = then median income would have risen along w average income, but it did not. But in superstar markets, higher likelihood of extreme events.
*Superstar Economy* = *"long-tail" economy* (low barriers to entry, but more inequality) "Winner-Take-All" market = unstable / asymmetrical *Power Law:* *small # people = reap disproportionate share of sales* (80/20 Rule: 20% of participants get 80% of gains) - *fat tail* = likelihood of extreme events = much greater than in normal curve
How do the authors conclude the book? - *Our values = will matter more than ever* - Will we widely disseminate info or tightly control? - Will prosperity be shared? - Will we build vibrant relat/communities? - Will everyone have opp. to discover/create/enjoy best of life? - *Technology = NOT destiny. WE SHAPE our destiny*
*Techn = creates possibilities & potential* but *ultimately, future* we get = *will depend on choices we make* - *Techn. we create* = provide more *power to change world.* & *greater responsibility* - As more work done by machines = ppl. can spend more time on other activities
How do the authors compare traditional markets (ex. the "90%") with today's markets? Each time a market becomes more digital = "winner-take-all" economics become more compelling (*"Winner-Take-All:"* *compensation* = det. by *RELATIVE performance*) (*"Traditional Market:" revenue* = closely tracked by *ABSOLUTE performance*) *Ppl = will NOT spend time/effort on tenth-best product when can get the best* = *consumers care about relative performance* (even a small diff. can lead to huge diff in earnings for company)
*Traditional* Market = *someone who is 90% as skilled / works 90% as hard = creates 90% as much value & can earn 90% as much money* = ABSOLUTE performance *Winner-Take-All* = ex.) programmer who *writes SLIGHTLY BETTER map app = might COMPLETELY DOMINATE market* = RELATIVE performance ex. Waze vs every other GPS map app = *little demand for tenth best map app even if almost as good* why use different social media site when can use Facebook (the best)?
Note the author's point in the section starting, "another school of thought."
*True work of innovation = not coming up w something big & new* instead, *RECOMBINING things that ALREADY EXIST* ex.) PCR = technique for replicating DNA was created by recombining well-understood techniques in Bio-Chem to generate a new one ex.) "*New Growth Theory:*" take resources & rearrange them in ways that make them more valuable (possibilities = don't "add up," they "multiply") ex.) *Meta-Ideas:* how to support production & transmission of other ideas
Review the section on economic theory and historical evidence for technological unemployment. 3. *SEVERE INEQUALITY* *Techn = created BOTH winners/losers via skills-biased tech change, capital-biased tech change, proliferation of superstars in winner-take-all markets* = *reduced demand for some types work/skills* In extreme winner-take-all markets = *equilibrium wage might be zero* ex. might offer to sing "Hello" for free, but ppl would still pay to here Adele sing it & she could make digital copies that compete w us (rational ppl. would look for new job) *near-zero wage = NOT living wage*
- *There is a FLOOR on how low wages for human labor can go* = *floor --> unemployment* (ppl. who want work but unable find jobs) *if a worker's skills not profitable or needed = that worker will go unemployed indefinitely* *Techn = can create inequality & unemployment* but...DATA: *for 200 yrs = techn. boosted productivity enormously* - employment grew alongside productivity until end of 20th Cent. (therefore: productiv. not always destroys jobs) HOWEVER *recently: job growth = unlinked from productiv. in late '90s* (which history to believe? past 200 yrs ending in '90s or past 15 yrs?)
Review the section on economic theory and historical evidence for technological unemployment. 1. *THEORY:* *If techn --> efficient use of labor = does NOT auto. lead to reduced demand for labor* - *Lower costs* --> to *lower prices* for goods & -*Lower prices* --> *greater demand* for goods & - Lead to *increase in demand for labor* (whether this happens = depends on *elasticity of demand* - % increase in quantity demanded for each % decline in price) ex.) cutting price of artificial light 1/2 = did not double amount of light consumers demanded, so revenues = fell as lighting became more efficient (*demand = inelastic = insensitive to price decline*) *INELASTIC DEMAND* ag & manufacturing = experienced falling employment as became more efficient
- Lower prices & improved quality of outputs = did not lead to enough increased demand to offset improved productivity BUT - When *demand = ELASTIC* = *greater productivity --> enough of increase in demand = more labor --> employment* Core Econ Argument = *techn. unemployment = impossible* (money is spent not just on buying more existing goods, but also on newly invented products/services) - increase in productivity = matched by identical increase in demand to keep everyone as busy as before (*Keynes Disagrees:*) - long run = demand would NOT be perfectly inelastic - even lower prices = not mean we'd consume ever more...instead *we'd become SATIATED / CONSUME LESS* ---> *reduction in working hours / less hrs needed* Ppl. = working less becuz satiated. Econ. problem of scarcity = replaced by what to do w abundant wealth/copious leisure *Goal of future = full unemployment so can play*
What do the authors note about future careers and information work? *Future* = *more careers will NOT be in pure information work* (desk work). - Instead = *will include moving through & interacting w physical world* (because *computers remain comparatively weak here* even as get stronger at many cognitive tasks) - *Far from cracking Moravec's Paradox* (took robot 24 min to fold each towel) - indicates: cooks / gardeners /repairmen / dentists / etc. = will not be replaced by machines in short term
- all these professions = *involve lots of sensorimotor work* & *ideation / large-frame pattern recognition / complex cmn* = but they may become subject to more human competition. As ppl. who previously did mid-skill knowl. work get automated = may start going after jobs lower on skills & wage ladder *Don't predict robots will acquire general skills of ideation/large-frame pattern recogn/highly complex cmn/solve Moravec's Paradox any time soon* Ppl = need be more adaptable/flexible in career aspirations. *seize new opp. where machines complement & augment human capabilities*
what do the authors claim about GDP growth and its causes
-In the US, the rate of GDP growth per person has averaged 1.9 percent per years going back to the early 1800s. This was enough to double living standars every 36 years, quadrupling them over the course of a typical lifetime. -If GDP of the US grows just 1% faster each year than currently projected, Americans would be five trillion dollars richer by 2033. causes: part of it comes from using more resources but most of it comes from increases in our ability to get more output from the given level of input--in other words, increases in productivity. *productivity growth comes from innovations in technology and techniques of production
what are the four categories of intangible assets?
-intellectual property: patent and copy rights -organizational capital: new business processes, techniques of production, organizational forms, and business models. -user-generated content:users of FB, Youtube, Twitter, Insta, pitnerest, etc. not only consumer this free content and gain the consumer surplus but also produce more of the content -human capital: the many years that we spend in schools learning reading, writing, arithmetic skills as well as other learning that happens on the job and on our own makes us more productive and in some cases in intrinsically rewarding. it is also a contribution to the nation's capital stock.
be familiar with the six policy proposals the authors make. Be particularly familiar with the authors education proposals
-teach the children well: when tech advances too quickly to for education to keep up, inequality generally rises. We can change the way we deliver education by putting to work digital technologies that have been developed over the past decade or two. -restart startups: entrepreneurship is the best way to create jobs and opportunity -make more matches: we must find ways to reduce the friction and search costs that make it unnecessarily difficult to match people up with jobs. support our scientists: US federal government support for basic academic research started to fall in 2005. this is cause for concern because economics teaches that basic research has large beneficial externalities. -upgrade infrastructure: government should be involved in building and maintaining infrastructure--streets and highways, bridges, ports, dams, airports and air traffic control systems, etc. b/c its subject to positive externalities -since we must tax, tax wisely: taxing something discourages its production. usually considered bad but it doesnt have to be since we can tax things we want less of.
What constitutes the Morris social development concept? *Social Development:* a group's *ability to master its physical & intellectual environ. to get things done*
1. *Energy Capture:* (per-person calories obtained from environ / home & commerce / industry & ag / transportation) 2. *Organization:* (size of largest city) 3. *War-Making Capacity:* (# of troops / power & speed of weapons / logistical capabilities) 4. *Information Technology:* (sophist. of avail. tools for sharing & processing info / extent of their use) *Social Development* = *sum of these 4 numbers* 14,000 BCE - 2,000 CE
What are the 3 key characteristics of the second machine age?
1. *Exponential* 2. *Digital* 3. *Combinational*
What are the 2 related weaknesses of complex system? - our species = now has power to destroy itself. - techn = reaching hands of more ppl. as techn becomes more powerful/cheaper/ubiquitous - computer programs = can self-replicate - physical limits on how much damage indiv/group can do = less constrained - Techn = could lead to loss of freedom, monitoring, inform flow control, unexpected side effects (gaming addition, digital distractions)
1. It is *subject to seeing minor initial flaws cascade via unpredictable sequence into something much larger / more damaging* ex. Three Mile Island: nuclear plant blackout affected 45million ppl. 2. *Complex, tightly coupled systems make tempting targets for spies, criminals, & those who seek to wreak havoc* ex. Stuxnet computer worm: hobbled Iranian nuclear facility by spreading PC to PC & hopping to other machines
What are the "three broad conclusions" the authors have reached in their research?
1. We're living in a time of *astonishing progress w digital techn* (building blocks in place for digital tech. to be as important / transformational to society & economy as steam engine = will be 2nd Machine Age) 2. *Transformations brought* about *by digital techn* = *will be profoundly beneficial ones* (better: cuz can increase variety & volume of our consumption) (Techn = bring us more choice and freedom / Digitization = improving phys. world) 3. *Digitization = going to bring challenges* (bring economic disruption = computers replacing people)
Why are winner-take-all markets more common now? 1. *Digitalization = creates "winner-take-all" markets* - cuz: *capacity constraints = irrelevant* w digital goods ex. single producer = can fill demand of millions of customers now (Jenna Marbles, 1 youtube video --> seen by 1Billion ---> she makes $millions) *every digital app developer* = *almost autom. micro-multinational, reaching global aud. fast*
2. *Digital goods = enormous economies of scale* - once fixed costs = covered --> each marginal unit produced costs very little to deliver "Winner-Take-All" = *boosted by techn. improvements in telecomm. & transportation = also expand market indiv. & companies can reach* Top performers = opp. to win more customers. While next best = face harsh competition from all directions (can no longer count on customer ignorance / geographic barriers to protect margins) ex. Hard to compete against Amazon ex. online retailers = make it easy to rank order by customer rankings / filter results by most pop.
Note the author's description of Khan Academy *Khan Academy:* started as series of online doodles / Youtube video lectures. - online edu. material free for all - more than 4,100 videos no more than few min long - range: arithmetic to calc to physics to art history - viewed more than 250 mill times - students = can tackle more than 1 Billions auto gen. problems - originally aimed at primary-school children, now applies to higher edu *Massive Online Open Courses* (MOOC)
2011: Sebastian Thrun = top artificial intell. researcher taught grad-level A1 course as MOOC (free, online. Not just for students at Stanford) - Test results from his course found that over 400 ppl. in the world outperformed the top Stanford student = *hope MOOCs eventually provide lower-cost alt. to traditional college edu*
Why are winner-take-all markets more common now? ex. when users prefer products/services that other ppl. flock too (If most of your friends use Facebook, it will be more attractive to you and you'll flock to it & it becomes more valuable to friend group) ex. bigger the network of users = will tend to attract more developers / encourage app develp. to invest more in that platform (*total # users on given platform = influences app developers*)
3. *Increased importance of networks* (Internet / credit card netwrks) & *interoperable products* (computer components) ---> create "winner-take-all" markets Networks = *create "demand side economies of scale"* (aka: network effects) *Your benefits from buying one or other = will be affected by # of other users who buy same product* ex. when Apple = strong, buyers will want to buy into that platform ---> attracts more developers
What happened to American household income from late 1990s to early 2010s? *1999:* *peak year* for *real income of median* American household ($54,932 but then started falling) *2011:* income median =*fallen 10%* to $50,054 - wages of unskilled workers = trended downward *2012:* *over half total income* in US *went to top 10% of Americans* - *top 1% = earned over 22%* of income *Life Expectancy:* fell for some groups Ave. American White Woman w/o high school degree = life expect. of (*2008:*) *73.5yrs* (*1990:*) 78.5yrs* (fell by 3 yrs for men too)
And what happened 1983-2009 w the "bottom 80%?" 1983-2009: Americans = *vastly wealthier overall & total value of their assets = increased* *bottom 80% of income distribution = saw net DECREASE in their wealth* Top 20% = got more than 100% of the increase Some additional wealth = shifted from bottom 80% to the top Top 5% = got 80% of nation's wealth increase Top 1% = got over 50% of nation's wealth increase
Review the section on economic theory and historical evidence for technological unemployment. 2. *RAPID CHANGE:* *Inability of our skills/organiz/institutions to keep pace w techn. change* - when tech elimin. one job type = those workers must develop new skills / find new jobs *Optimistic View:* this is temporary, eventually econ. = find equilibrium & full employment restored as ppl. invent new businesses & workforce adapts human capital
BUT: what if this process takes a decade & techn has changed again? - *workers = could end up permanently unemployed* - accelerating tech. change = lead to widening gaps / ^ possibilities for techn. unemployment *Faster techn. progress --> may bring greater wealth & longer lifespans* BUT *requires faster adjustments* by ppl/institutions
Note the section on China and manufacturing employment. - *integration of global markets --> enormous dislocations (labor markets)* American manuf. = expected to shift production overseas where costs = lower *China = responsible for 1/4 of decline in US manufacturing employment* BUT manuf. employment in China = has fallen too due to automation (more output, fewer workers)
Businesses = *can identify & hire workers w skills they need, ANYWHERE in the world* - If worker in *China* = can *do same work as American* then *"law of one price"* = *demands they earn essentially same wages* = *good new for Chinese worker & overall econ. efficiency* BUT *bad news for American worker facing low-cost competition*
Note the example near the end of the chapter on call center automation.
Interactive Voice-Response systems = automating jobs in call centers = this can *disproportionately affect low-cost workers in places like India/Philippines* *Intelligent & flexible machines (not humans) = MOST cost-effective source for 'labor'* - Types of tasks off-shored = *routine / well-structured tasks* = aka *easiest to automate* *Off-shored jobs --> on way to automation*
Note the issue of digital content coming freely
Many people = *willing to devote time to produce free online content* ex.) Wikipedia & Siri (improves self w user-gen data) - *User-Generated content* = *created free of charge* 6/10 most pop. content sites in world = user-generated
What is the "technological singularity?" *Technological Singularity:* an intellectual transition where *humans & computers could merge seamlessly* - humans = dont fight w machines, we join them (utopian view) - we *upload our brains into the cloud & become part of techn. singularity* "soon create intelligence greater than our own"
Progress towards singularity = driven by Moore's Law - accumulated doubling = eventually yield a computer w more processing/storage capacity than human brain = things will become highly unpredictable (*machines = could become self-aware & humans+computers could merge seamlessly*) could occur by 2045. Techn. not built same way nature built humans. Human-like skills/abilities = artificial resemblance.
Figure 9.3 and accompany text describe what story? Who is capturing the value of new productivity? *Owners of physical capital* = *capturing the value of new productivity* (productivity = growing, but labor as whole not capturing this value) ex. spending on software/capital equip = soars payrolls = flat (soaring wages for small # superstars & CEO's income = due to their control of capital, NOT labor)
Shows that in *past decade: the relatively consistent division bet. : - shares of income going to labor & physical capital* = *coming to an end* - share of *GDP going to labor = declined* (*likely due to techn. of info. age*) & global phenom. *Fewer ppl. working & wages for working = lower* *Labor Compensation + Productivity* = used to rise in tandem, now = *growing gap * *Share of national income : Capital = growing at expense of labor*
What do the authors state about "superstars" vs "second bests?" ex. only 4% of software developers have made over $1million 3/4 = made less than $30K - only handful of actors make millions, rest struggle
Super-duper stars = pulling away from superstars *When improvements in digital techn. make* it more *attractive to digitize* something, *SUPERSTARS = see INCOME BOOST * *SECOND-BESTS* = *have HARD TIME COMPETING*
what is the authors' point about the example of music and economic numbers?
The cost of things like music has gone down since it has become digital. leads to different economics and some special measurement problems.
Note the role of the British Empire described by Sugata Mitra in text. (gave computers to poor kids in developing nations & they learned on their own) *British Empire* = where the kind of learning we do in schools originated from - They = created global computer (bureaucratic administrative machine) made up of people - To run it = need lots of people - Created the "school" to produce these people who would become parts of the bureaucratic admin. machine - Must have/be able to: *good handwriting* (hand written data), *read, math in their head* (mult/div/add/subtr) - Must be so identical they could pick someone from Australia & ship then to Canada & he'd be functional
This edu. system = good for Victorian England, but design does not work well for our time Today = schools are producing identical ppl. for a "machine" that no longer exists - today, clerks are computers - don't need to write by hand or do math in head = computers do it - DO need to be able to *read & read discerningly* ex. ppl. now need to *acquire / demonstrate skills of ideation / broad-frame pattern recognition / complex cmn* "Self-Organizing Learning Environment" (SOLE) = teaching children the *skills* that will *give them advantages over digital labor.* ex.) Montessori Schools (self-directed learning)
Who are the three sets of winners described by the authors? (of structural economic changes) In each group = digital techn. *increase economic payoff to winners* while *others = become less essential / less well rewarded* *Overall gains to winners = larger than total losses for everyone else*
Those who have *accumulated significant quantities of right capital assets* 1. *Non-Human Capital* (equipt, structures, intellec prop, financial assets) 2. *Human Capital* (training, edu, experience, skills) asset = can generate stream of income (favored by skills-biased tech. change) 3. *Superstars:**special talents - or luck*
what do the authors argue about free products? (note here the example of free newspapers vs. those with a price)
replacing a paid newspaper with an equivalent free new service would decrease GDP even though it increased consumer surplus.
what explains the lag that exists between the introduction of a technology and productivity benefits?
the slow start and subsequent acceleration of productivity growth in the electricity era matches well with the speed-up that began in the 1990s. the key to understanding this pattern is the realization that GPTs always need complements, coming up with this can take years or even decades.