Life After Robots: The Jobs Report

The giddy technology boosters of Wired magazine can’t wait for our robotic future, when we will give thanks to the robots for taking our old jobs away so that we can do more interesting and better paying work: “Wow, now that robots are doing my old job, my new job is much more fun and pays more!”

Unfortunately, this optimistic wish-thinking shows an embarrassing ignorance of economic history, including recent history.
For the reality is that workers who lose their jobs to technology tend to end up unemployed or employed at vastly lower wages. Society as a whole may end up better, but the displaced workers generally do not, even if their children do. See Rust Belt, USA. Or Textile Workers, England.

I believe the premise is correct: software and robots will indeed transform the workforce and the professions. Quite possibly 70% of job positions today will no longer exist at the end of the century. According to other estimates, the number may be closer to 45% of jobs taken away by robots.

And whereas it took decades to render farming jobs obsolete with mechanical tractors and farm equipment, software and robotics are evolving so quickly that they are likely to change what skills and jobs are needed at a faster rate.

Unfortunately, evidence to date suggests that workers cannot retrain so easily, even when they’re motivated and interested in changing careers. How many unemployed middle-aged factory workers, truck drivers, warehouse workers, cashiers, clerks, administrative assistants and others succeed in going back to college and learning how to write and use advanced software and operate high end robotics?

We know the answer: not many. Large numbers of displaced workers can expect to spend extended periods of time on the unemployment rolls, only to discover that companies don’t want to hire them because they’ve been out of work. If they are lucky enough to get hired, they may find a new job at lower pay, often part-time, with fewer benefits and a lower title.

Yet there are still at least three good reasons for optimism.

  1. Technology costs are dropping rapidly.
  2. Workers may be becoming more adaptable (though the evidence is not yet in).
  3. And 70% job loss may not be as bad as it sounds.

Technology costs are declining rapidly, so that the same money can buy better, and more, and newer. Despite stagnating wages, and outright declines for those whose jobs have been replaced by technology, flat income can in fact buy more.

A tablet computer today costs a few hundred dollars and provides far more functionality than a $2,000 laptop did in twenty years ago. Taking inflation into consideration, the same salary can buy a vastly superior computer for on the order of 1/10th the cost.

Could we be headed toward a future where most people live on minimum wage jobs, or on unemployment checks, and yet they live better than people at any other time in history? Where a small number of fabulously wealthy people live like feudal lords? This is the future imagined in The Lights in the Tunnel.

That brings us to the second possible reason for optimism. Perhaps workers today are becoming a different breed, whose skills can adapt more easily.

Switching from a semi-skilled job on a factory floor to a computer operator or programmer for new robotics may be difficult. Switching from a computer operator for one type of robotics to a computer operator for the new generation of robotics may be easier. In other words, certain kinds of changes — from physical labor to service — may be a one-time switch, whereas jobs within a given category may be easier.

People may be more adaptable, and the adaptation may already be underway for the younger generations. The cycle of rapidly upgraded computers and telecommunications may be rewiring our brains for the economic future.

If technology continues to lower costs, and if people are indeed becoming more adaptable, and hence less vulnerable to technological unemployment, maybe the painful downside of the robotic revolution will be reduced, even without government intervention.

And in case you’re worried by a 45% or 70% job turnover from robotics and computers, that number is actually small compared to the normal turnover from non-technological causes. Around 15% of all jobs are lost without replacement in a given year. In that case, robotics will bring but a blip in the digital photo of a bucket.

Even if you believe the problem isn’t so big, there will be many real people who lose out in the change, who see sharp drops in their standard of living, and never look back with happiness on when they were freed from their previous work and given new, higher paying and more meaningful work to replace it. Overall, the trend of increasing human prosperity is likely to continue. Robotics will be part of it.…

Read more

Silicon Valley Housing Crunch

Rents and real estate prices have gone crazy in Silicon Valley and San Francisco. What’s going on? Is it sustainable or a high tech bubble all over again?

Housing Crunch
San Francisco residents are complaining about the high tech companies — Google, Facebook, Twitter, Apple — that have contributed to a leap in rental prices, with the median one bedroom apartment now costing $2,795, and considerably higher in some neighborhoods.

Meanwhile condos and houses throughout the Bay Area are selling with multiple offers at prices above list, and above the peak pricesof the housing bubble. All-cash offers are common, and real estate agents recommend a 30% down-payment to even get considered.

Part of the housing crunch in San Fran proper is caused by commuters to high tech companies, enabled by the controversial luxury buses with leather seats and wifi that transport young technologists to the tech titans of Silicon Valley.

Commercial real estate is equally affected. With the influx of high salaried workers, expensive boutique shops and fancy restaurants are replacing more modest establishments, driven out by the high rents.

The Great Inversion
Isn’t it just gentrification? Same story happening everywhere. Yes, and no. San Francisco prices have been rising for a long time, and the merely upper middle class are being priced out by the even better off.

A demographic inversion is continuing, with the formerly decaying and crime-ridden inner cities getting richer and the suburbs becoming the new home of the immigrants, as in European cities.

It Will Continue
For multiple reasons, the trend is likely to continue.

The population is growing. That puts upward pressure on prices close to where jobs are.

Income inequality is increasing. There are no signs yet of a change to the decades long trend of rising inequality. As the rich get richer, the richest will spend more of their money on their homes, driving up prices in the most desirable neighborhoods.

Urban planning and legislation suppresses the supply of housing. Places like the Bay Area have restrictive development and zoning policies, and perverse tax incentives, which restrict supply. That inflates prices to the benefit of existing homeowners, reducing turnover.

Geography limits expansion. San Francisco is surrounded by water on three sides. There’s no room to expand, unlike in some cities surrounded by developable land.

Finally, the technology revolution inaugurated by the Silicon Age is not slowing down. The Bay Area remains the center of high tech investment, and as long as computing continues to transform more areas of life, more tech billionaires and multi-millionaires will bid up the prices.

If it were only Silicon Valley, the rest of the world could shrug and say, Who cares? But the patterns of rising prices can be found in many urban centers throughout the US and the world, regardless of the dominant industry. Where the local economy is vibrant, housing prices are rising in many inner city and inner suburban neighborhoods.

Multiple of the trends favoring rising prices can be found in other urban areas with a solid economy. The geographic constraints can be found in most coastal cities or cities bordered by lakes or mountains. Affluent and liberal areas everywhere tend to pass laws to restrict development. Only the technology trend distinguishes Silicon Valley, and even then, several other urban areas have a growing base of high tech jobs.

I’m not very satisfied with what I’ve read about the revitalization of cities, gentrification, and housing trends generally. I do recommend reading The Great Inversion for part of the story.…

Read more

Where’s the Money in Big Data?

So asks NY Times writer James Glanz in his recent article, “Is Big Data an Economic Big Dud?”

The short answer to his question is no. Now for the slightly longer answer.
What’s So Great About Big Data?
It may not strike you as immediately obvious why more data should be a good thing. Does the world need more YouTube videos of cats, or Facebook posts, or tracking data of your web clicks? How does that create economic value?

Simplistically, of course, the companies collecting the data, and convincing users to generate it, are doing pretty well. Facebook stock has risen back to its already phenomenal IPO valuation. YouTube turns out to have been a bargain for Google at only a $1 billion price tag. And online advertising continues to grow at double digits.

Looking deeper, the data craze builds on a more fundamental observation, brought into the ranks of corporations by the likes of Peter Drucker in the 1940s and 50s: business decisions ought to be based on data. Or at least not to contradict the data.

In the sciences, the turn to data and quantification and mathematics has been on a centuries long march that has recently penetrated into the social sciences, including psychology and marketing. One way to look at the Big Data hoopla is simply as the logical extension of this: if you have data, and you can improve business decisions — like the choice of advertising to show to users — you can make more money.

The premise is simple enough, and the results are indisputable in fields like advertising.

What’s Not So Great
Small businesses can’t benefit from data in the same way that larger ones can, because they don’t have enough customers of their own to extract trends, and improving their decisions on average may not help if they only have one store location to average across.

Big Data will benefit big businesses first. From big box stores to Big Data to mega-airline mergers, the trend is toward bigger. That’s not so great if you care about the idiosyncrasy of the small, if you like the autonomy of the mom and pop store.

As a result, small businesses have little choice but to rely on big companies to provide the data they can use. They must advertise using the aggregators, play the game of rankings and search engines and social media, and find a way to stand out in the noise.

They depend on the Amazons and Apples of the world to be their distribution channel, knowing that larger companies can copy their ideas and produce them at larger scale and lower cost if they’re successful.

Drowning in Data
Data has become too easy to collect and store. Our memories are fallible and impermanent. Our photos on Instagram reveal more than we can recall.

With inventions like Google Glass, a future where every human’s every waking moment is recorded in high fidelity 3D video is not so far-fetched.

Companies, too, can collect more data than ever before, and they can actually search through it for trends, and adjust their product strategies, advertising, and pricing based on what they find.

It’s pretty obvious that there’s money to be made from not just collecting the data, but analyzing it and acting on it as close to instantaneously as possible. In economic terms, more accurate data can lead to greater efficiency, by detecting mis-pricing and missed opportunities.

The Big Data industry has risen up to provide the technology tools so that every company, in every industry, can harness, store, search, mine, learn from, and act on the data that can help it to perform better.

Why Big Data Might Not Pan Out
With such clear benefits, why would anyone doubt the economic contribution? Here are a few legitimate concerns.

Maybe Big Data will not primarily create new value, but mainly shift the value to those who collect and use the data better than the competition, in the process enriching the companies that develop the tools.

Maybe more data will face diminishing returns very quickly, rather than creating new sources of value. Smoothing out some suboptimal pricing, and luring in a few extra shoppers with better targeted offers, could provide a marginal benefit, not a transformative and sustainable one.

Big Data could turn into an arms race, similar to high frequency trading. Or, well, literal arms races. Big Data could be a more powerful weapon, and the best armed companies will destroy the competition. Free market zealots will hail the Schumpeterian “creative destruction”.

In the past, the destruction has tended to create more opportunities than it has lost, over the long term, despite the undeniably negative effects on those immediately impacted, like the mill workers unemployed, or the auto workers whose salaries have halved.

With the rise in income inequality to levels last seen at in the gilded age, and with the “hollowing out” of the global workforce, the past may not be as reliable a guide.

Even if one of these skeptical scenarios come to pass, Big Data will not be a Big Dud, at least not for all.

In a more optimistic outcome, which I consider to be more likely, Big Data will be a net win, keeping the energy of economic growth thrumming at a pace to lift hundreds of millions more around the world out of poverty in the coming decades, bringing a higher standard of living, longer lives, and less misery to the world than at any time in history. Big Data can hardly take credit as the only or even the main cause. It will be but one minor contributing factor. But I believe it will contribute positively.

Ignore the data — like ignoring the facts — at your own peril.…

Read more

Real and Virtual Violence

As the US contemplates a military strike against Syria, it’s easy to lament the violent state of the world. But real-world violence has declined remarkably, as detailed in Steven Pinker’s book.

Instead, violence has become virtual– moving into video games and movies, like the summer’s blockbusters. Previously, I argued that (virtual) violence has actually increased with technology, because it’s far easier and less risky to kill video game villains than to pick a real physical fight.

Now I want to build on that point, and join a couple of threads in my articles. Violence and combat capability have achieved a state that I call the perfection paradox, a state where military technology has so perfected the art of killing that physical martial skills and muscled bodies have become essentially irrelevant.

Summer action movie heroes dodge bullets, and fight with fantasy martial arts moves in musculature physique sculpted by steroids and a personal trainer. Meanwhile we all know that they’re no match for a real-world, malnourished child soldier with a third-hand gun and a functioning trigger finger.

Old style martial arts training is for entertainment, and serves essentially no purpose in real combat. Overweight suburban parents send their children to kung fu class to make up for canceled phys ed classes. Winning enough safe fights might help earn college admission. It won’t result in any advantage in an actual war.

Meanwhile, dedicated athletes head to the gym several times a week to cultivate body builder muscles at the gym. But it’s all for show, for sex appeal. They won’t be fighting with those muscles.

That’s not to say that exercise is irrelevant. In rich countries like the US, we don’t get enough exercise.

For useful muscles, look to the migrant laborers working in the agricultural industry, or construction workers. Those aren’t muscles for war, but muscles for survival.

War has always been driven by technology, from flint knives to spears to combat horses, and then to gunpowder, armored vehicles, missiles and drones. Chemical weapons are among the cast of real world weapons of the day.

Strength training and martial arts classes won’t protect us. They’re for show only, for showing off and for show business in video games and movies.…

Read more

Cities of the Future

Well-heeled liberals tend to assume that everyone should live in cities, and the world would be a better place for all. People would consume less energy for transportation and temperature control. They’d get more exercise from walking and taking public transit, extending their lives. Quaint Jane Jacobs-style neighborhoods would improve people’s daily routines with friendly local businesses and amenities.

All of that’s true. But here’s why it could go a different way.

The City Is Dead
Given technology trends, there will soon be fewer reasons for cities to exist.

  • Video conferencing improves so much that physical proximity becomes irrelevant — for socializing as well as working
  • 3D nano-printers one day can create any food near instantaneously, driving most restaurants out of business, because we’ll all be able to Jetson our dinners on demand. Assuming that virtual reality doesn’t render the physicality of food irrelevant first.
  • Online shopping continues along its trend toward beating the real thing.
  • Online dating utterly replaces the in-person variety, as the recent marriage stats suggest.

If all these come to pass, what reason will be left for living in a city? It won’t matter where anyone lives.

Long Live the City
And yet, some of the benefits of cities will remain.

  • Energy efficiency
  • Greener living
  • Lower cost delivery of goods
  • Virtual reality may one day provide an adequate experience of nature

In an especially science fictional future, people might end up living in Matrix movie style pods, plugged into virtual reality, living at a density to make Tokyo seem spacious.

What do you think will happen?…

Read more

Can Computers Compose a Hit Music Single?

Computers can beat us at chess. They can beat us at Jeopardy-style trivia. They can out-calculate us and make fewer mistakes on rote problems.

Can computers write better music? What would it take for them to compose a #1 hit music single?

For ideas, I finally got around to reading Raymond Kurzweil’s book The Singularity Is Near, a touchstone of futuristic thinking.

Consciousness in the Cloud
In Kurzweil’s vision, the Singularity is the time when computers surpass people in intelligence — and not just intelligence, but effectively all ways. He believes that people will choose to merge with computers, and replace their biological bodies with other physical forms, or with pure software, uploading their consciousnesses into what we now call “the cloud” of computer systems distributed around the world.

Naturally, in Kurzweil’s vision, computers will be better than humans at art and music too.

But what would it take?

The Computer Music Challenge
Suppose that IBM were to build a supercomputer called Golden Ear, whose mission is to top the global popular music charts, beating out the Los Angeles music industry as well as the music industries in Korea, Mumbai, Tokyo and everywhere else.

IBM already developed Blue Gene for chess and Watson for Jeopardy. Now Watson is taking on other challengs, like medical diagnosis.

Is Golden Ear the successor to Blue Gene or Watson? Or a new class of computer?

Designing a Musical Computer
How can a computer know what melody or chorus sounds good to people?

By analyzing trends in music, popular culture, news, celebrities, vernacular language. Then choosing sounds and lyrics that capture the moment.

Consider that mobile phone apps like Shazam can already recognize music from a few notes and retrieve the title, artist, lyrics and more. IBM’s Watson can mine vast troves of text and guess about puns and trivia.

Is it a coincidence that the Lady Gaga hit “Poker Face”, released on Sept 2008, arrived not long after the peak of the poker mania? Look at the plot of Google searches for poker and you’ll find that the mania peaked in Jan 2007, close to the height of the housing bubble. “Poker Face” the song consists of upbeat music and that catch phrase, repeated often. The song title itself peaked as a search term in April 2009, 7 months after the album’s debut.

Could a computer have identified the poker trend, selected a catchy phrase timed to the bursting poker bubble, and produced and launched a song on iTunes earlier than Lady Gaga did? Could such a computer catch the next trend?

Would the music industry pay royally for computer software that wrote songs for a stable of bankable performers and generated a higher hit rate than human composers?

The Golden Ear Challenge
Previous efforts at computerized music composition have not met with much success. That’s an understatement, as a matter of fact.

But computers have improved rapidly, and continue their exponential improvements. As they did with chess and trivia, computers could overtake humans in this new challenge.

Kurzweil would not be surprised. The Singularity may be near for artificial intelligence. Is the #1 Single nearer?…

Read more

A Modest Proposal for Silicon Valley Housing

A modest proposal to fix the Bay Area housing crunch: Fill in the southern part of San Francisco Bay with landfill and build a 21st century city of earthquake resistant glass.

All of the land from the Dumbarton Bridge south could be converted from salt marshes and foul smelling sludge to a glorious city of the future. That could solve the housing shortage and bring prices back to earth, simultaneously allowing Silicon Valley to expand and grow still more.

It’s no secret that Silicon Valley has the highest real estate prices in the US. The reasons are many, including high salaries, pleasant climate, beautiful geography, jobs — and a dire lack of houses.

Yes, Silicon Valley real estate prices are the highest in the nation for many reasons. And all of them could go away if governments just let developers build enough more, and limited the ability of existing property owners to block what other people can do on their own land.

Local communities exercise near veto rights over development near them, and city planners and developers find it difficult to win many small battles over tiny plots of land. Broad expanses of land abound in the ridge of hills and mountains between the bay and the ocean. But those are widely considered beautiful, and it’s near inconceivable that such stunning parkland will be developed anytime soon.

Meanwhile, the southern part of the bay is not generally considered beautiful. It’s filled with famously smelly salt marshes that serve only modest ecological value. Their benefits could readily be offset with eco-friendly preservation somewhere else.

Even better, the water comes within walking distance of Facebook, Google, Yahoo, Cisco and Tesla. It’s a prime location, much more so than San Francisco, so many miles further north, away from the high tech center of the Bay Area.

A new city, literally in the bay, would provide expansion area and take pressure off of housing prices. Better yet, an inchoate city would allow the vaunted technology visionaries of Silicon Valley to experiment in urban design, engineering the public transit, housing, and earthquake resistance of the future, supporting their own high tech economy and showing an example for the world.

And no human infants need be eaten or sold.

Why not? Please share your reasons.…

Read more