Irving Wladawsky-Berger

Skills and Jobs in the Digital Economy

Few subjects are as important, - or as challenging to predict, - as the future of jobs in our emerging digital economy. While the US unemployment rate continues to improve, - finishing 2014 at 5.6% or 8.7 million people, -  almost one third of the unemployed have been jobless for over 27 weeks.  The total number unemployed or underemployed, - what economists call the U6 unemployment rate, - stands at 11.2% or 17.4 million people.  And, the employment to population ratio remains at under 60%, the lowest such percentage since the 1970s.  “The economic challenge of the future will not be producing enough.  It will be providing enough good jobs,” wrote Harvard professor and former Treasury Secretary Larry Summers in a recent WSJ article. 

Technological revolutions are highly disruptive to economies and societies.  This was the case for much of the Industrial Revolution, as is the case today.  “The digital revolution has yet to fulfil its promise of higher productivity and better jobs,” said The Economist in a special report on Technology and the World Economy in its October 4th issue.  “The modern digital revolution - with its hallmarks of computer power, connectivity and data ubiquity - has brought iPhones and the internet.”  But, “it is disrupting and dividing the world of work on a scale not seen for more than a century.  Vast wealth is being created without many workers; and for all but an elite few, work no longer guarantees a rising income.” 

Middle class jobs have been in decline for the past few decades in the US and other advanced economies, - particularly since 2000.  And, the livelihood of significant additional workers is potentially threatened, as our increasingly smart machines continue to be applied to activities requiring cognitive capabilities that not long ago were viewed as the exclusive domain of human.  How will these relentless advances in technology and automation affect the balance between humans and machines in the workplace, and the skill composition of future jobs? 

This critical question was addressed in a recent paper, Racing With and Against the Machine: Changes in Occupational Skill Composition in an Era of Rapid Technological Advance, by MIT’s Frank MacCrory, George Westerman and Erik Brynjolfsson along with Yousef Alhammadi from the Masdar Institute in Abu Dhabi.  The paper analyzed the changes between 2006 and 2014 in the skill composition of 674 occupations using the US Government’s O*NET data base, - the most comprehensive data sets of occupational skill requirements.  The period from 2006 to 2014 saw the advent of several major digital innovations, including smart mobile devices, social media, big data and analytics, cloud computing and the Internet of Things.

As Big Data and AI Take Hold, What Will It Take to Be an Effective Executive?

Big data, powerful analytics and AI are everywhere.  After years of promise and hype, technology is now being applied to activities that not long ago were viewed as the exclusive domain of humans.  Our digital revolution had led to amazing applications, but also to considerable pain for many workers who’ve been experiencing declining employment and wages.  Mid-skill jobs have been particularly threatened.  Many of these jobs, - which include blue-collar production activities as well as information-based white-collar ones, - are based on well understood procedures that can be described by a set of rules that machines can then follow. 

But, what will be the impact of our increasingly intelligent machines on senior management positions?  In principle, such jobs deal with non-routine, cognitive tasks requiring high human skills, including expert problem solving, complex decision-making and sophisticated communications for which there are no rule-based solutions.  “As artificial intelligence takes hold, what will it take to be an effective executive?” asks a recent McKinsey article - Manager and Machine: The new leadership equation.  “What would it take for algorithms to take over the C-suite?  And what will be senior leaders’ most important contributions if they do?” 

After asking these questions to senior managers across a broad range of industries, McKinsey  concluded that two key things need to happen for technology to more deeply transform their jobs.  First, much still needs to be done to create the proper data sets that would enable intelligent computers to assist in decision-making.  Garbage in, garbage out applies as much to data analysis today as it has to computing in general since its early years.  Organizations must have a data-analytics strategy that cuts across internal informational silos and properly incorporates external information sources like social media. 

And most important, senior managers must learn to let go, something which is quite difficult because it runs counter to decades of organizational practices.  Given our rapidly rising oceans of data, the command-and-control approach to management, where information flows up the organization and decisions are made at high levels, would sink the senior executive teams.  As data science and AI permeate the organization, it’s important to delegate more autonomy to the business units that hopefully have the proper skills, the advanced tools and the necessary information to make better decisions on their own.

“Soft” AI Is Suddenly Everywhere

“Artificial intelligence is suddenly everywhere.  It’s still what the experts call soft A.I., but it is proliferating like mad.”  So starts an excellent Vanity Fair article, - Enthusiasts and Skeptics Debate Artificial Intelligence, - by author and radio host Kurt Andersen.  Artificial intelligence is indeed everywhere, but these days, the term is used in so many different ways that it’s almost like saying that computers are now everywhere.  It’s true, but so general a statement that we must probe a bit deeper to understand its implications, - starting with what is meant by soft AI, versus its counterpart, strong AI. 

Soft, weak or narrow AI is inspired by, but doesn’t aim to mimic the human brain.  These are generally statistically oriented, computational intelligence methods for addressing complex problems based on the analysis of vast amounts of information using powerful computers and sophisticated algorithms, whose results exhibit qualities we tend to associate with human intelligence.

Soft AI was behind Deep Blue, - IBM’s chess playing supercomputer which in 1997 won a celebrated chess match against then reigning champion Gary Kasparov, - as well as Watson, - IBM’s question-answering system which in 2011 won the Jeopardy! Challenge against the two best human Jeopardy! players.  And, as Andersen notes in his article, it’s why “We’re now accustomed to having conversations with computers: to refill a prescription, make a cable-TV-service appointment, cancel an airline reservation - or, when driving, to silently obey the instructions of the voice from the G.P.S.” 

This engineering-oriented AI is indeed everywhere, and being increasingly applied to activities requiring intelligence and cognitive capabilities that not long ago were viewed as the exclusive domain of humans.  AI-based tools are enhancing our own cognitive powers, helping us process vast amounts of information and make ever more complex decisions.

The Startup Version of the Wealth Gap

Entrepreneurship has never been easier, but entrepreneurship is on the decline.  I first heard this surprising paradox a few months ago in a talk by MITs’s Andy McAfee.  Digital technologies are inexpensive and ubiquitous, startups have access to all kind of cloud-based business services, and customers can now be easily reached and supported over mobile devices.  It should be easier than ever, - as books and articles keep reminding us, - to become an entrepreneur and start your own company.  But in fact, entrepreneurship has been in decline for years. 

Most everyone I’ve mentioned this paradox to, - business colleagues, investors, journalists, - doesn’t quite believe it.   But that’s what the data shows.  In a recent paper, University of Maryland economist John Haltiwanger and collaborators used business data from the US Census Bureau to calculate the annual startup rates over the past several decades, - i.e., the number of new firms in each year divided by the total number of firms.  The startup rate was 12.0% in the late 1980s, went down to 10.6% just before the 2007 Great Recession, and then fell sharply below 8%.  Such sharp declines add up over time.  In the late 1980s, 47% of all firms were 5 years or younger.  The percentage of young firms declined to 39% in the mid 2000s, and has since continued its downward trend. 

In another recent paper, economists Ian Hathaway and Robert Litan analyzed the same Census Bureau Data and showed that in addition to the continuing decline of new firm formations, failure rates have steadily increased for all companies under 16 years old, and they’ve been particularly high for early-stage firms.

Further evidence of this entrepreneurship decline is found in Where the Jobs Are: Entrepreneurship and the Soul of the American Economy, published in September of 2013 by John Dearie and Courtney Geduldig.  Their book is primarily focused on the close link between entrepreneurship and job creation.  It cites a 2009 study that shows that between 1980 and 2005, all net new job creation in the US came from young businesses less than 5 years old.  They conclude that as a result of the decline in entrepreneurship, “the nation’s job creation engine - new business formation - has been breaking down in recent years.”