Whether you're an aspiring lawyer, policeman or programmer, you should be aware that at some point — maybe a decade from now, maybe two, perhaps less — many jobs in those industries will be replaced by an algorithm.
That's what many economists predict and in some cases it looks like it will happen very soon. For instance, algorithms currently perform some tasks previously executed by paralegals, patent attorneys and contract lawyers. In Doha, Sao Paulo and Beijing, municipal governments use cheap sensors on pipes, pumps and other water infrastructure components to watch out for water leaks, a practice that has led some to speculate that fewer law enforcement workers will be needed on patrol once more sensors are deployed. Even programming — once the epitome of a safe-as-milk job in the 21st century — could be taken over by the bots as machine learning lets algorithms make and optimize design choices in coding.
All told, some 47% of U.S. employment is at risk of being automated over the next two decades, according to a 2013 study by Carl Benedikt Frey and Michael Osborne of the University of Oxford.
All of this means that if you're in college now, it's diabolically tricky to try to pick a career that you'll still be doing when you're 40, assuming that the world is still around in some form or another. While economists are a bit hazy when it comes to predicting new occupations that might replace the faltering old ones, they do offer some general guidelines about career development.
See also: 10 Jobs That Are Being Replaced By Machines
Basically, the idea is that you have to be able to outsmart a computer. "Look for things humans can do, but machines can't," says Erik Brynjolfsson, a professor at MIT and co-author (with Andrew McAfee) of The Second Machine Age. "Not routine information processing."
How can you outsmart the bots? Here are a few suggestions:

nited HomeCare Services home health aide Wendy Cerrato talks with Olga Socarras during a visit on Jan. 6, 2010 in Miami, Florida.
Image: Joe Raedle
One growth area is so-called "high touch" jobs like personal trainers, home health aides and wedding planners. These are jobs, in other words, that can't be outsourced to India or China. Rhys Grossman, a managing director of consultancy Russell Reynolds Associates, predicts that some of the new jobs created in the next few years will be along the lines of Joachin Phoenix's character in Her, who wrote letters on behalf of people who were too busy to do so themselves.
Matt McGraw, CEO and founder of The Lions, a tech recruiting firm, says high-touch extends to the sales profession. McGraw says as more consumers get access to data that was previously proprietary, a smaller percentage will need hand-holding, but those that do will be willing to pay for the value-add. "In 2034 instead of 100,000 real estate agents there might be 5,000," he says. "They'll be working with consumers who need high touch."
"I can only recommend that young people continue to gain the kind of cognitive and creative skills that give them a competitive edge over machines." That's the advice of Osborne, a co-author of the Oxford study. "In particular, and I may be biased, but occupations revolving around creative uses of data are likely to be resistant to automation for some time."
The key word here is "creative." If you're contemplating a career in which grunt work is the norm — like accounting, legal work and clerical jobs — your job may well be replaced by an algorithm by the time you're ready to hit the work force.
That's not say that those jobs will be completely automated. Picture, for instance, the self-checkout systems in the supermarket. They may be good for most transactions, but you need a human there when the system doesn't recognize an object or freezes up for one reason or another. Similarly, a software program may do a great job processing tax returns or legal papers, but a pair of human eyes is necessary to catch flaws that might get by a machine.
But being creative is not the same thing as being cognizant. In addition to being on hand in case the bots break down, Brynjolfsson says the economy will still need people to come up with new things, "make leaps into new areas" and synthesize information in new ways. In other words, starting up your own company will be less of a gamble and more of a logical career choice.
Leadership is another skill that can't be outsourced to an algorithm. For Brynjolfsson leadership means "motivating people, persuading, caring." Tuck Rickards, another managing director at Russell Reynolds Associates, agrees that leaders will be in great demand, but not middle managers. "My gut says the number of large established companies with middle management could go down," he says.
Instead, Rickards envisions hot-shot CEO types who might bounce from project to project rather than run a company per se. The reason? Technology like cloud computing and automation will let small staffs do work that previously required many hands. Rickards points out that we're seeing this already with companies like Instagram that sold for $700 million in 2012 and had just 13 employees at the time. Similarly, Snapchat, which has reportedly drawn a $3 billion bid from Facebook, has around 20 employees.

Basketball Player Deron Williams (R) and barber Marcos 'Reggae' Smith attend GQ Barber Shop Grand Opening with Fellow Barber at Barclays Center on Jan.10, 2014 in New York.
Image: Dimitrios Kambouris
There are some things that machines are great at — like analyzing legal documents — and some that they aren't. One such area is manual tasks that are easy for humans, like picking a dime up off the floor. "It turns out, that's incredibly hard if not impossible for robots," Brynjolfsson says.
Hence the expected viability of barbers, gardeners, cooks and other types of manual labor that would befuddle a bot.
The standard advice to someone embarking on a career is to do something you love and do it better than anyone else. Gazing ahead to the largely automated workplace of 2034, economists offer the same recommendation.
The journalism profession is an apt illustration of this phenomenon. We are now at the point where algorithms can compose basic news stories. However "You can't get an algorithm to do a Watergate," as Grossman says. Of course, to break a Watergate-type story you have to be very good at what you do, not to mention persistent. Those types of breakthroughs don't come at a job you're only lukewarm about. "It's important to pursue something you're passionate about," says Brynjolfsson. "it's hard to be really, really good at something."
The flip side of this is that — as we've seen with middle managers — those who are mediocre will have a harder time. However, employees should also consider the Long Tail. "There are global audiences for small niches now," Brynjolfsson says. For instance, if you're a journalist, you might consider specializing in a somewhat arcane topic — like green technology, for example — rather than trying to go toe-to-toe with The Washington Post and The New York Times on national political stories.
Workplace automation, fewer jobs and an increasingly winner-take-all society do not necessarily bode well for democracy. In 2013, the top 85 individuals in the world owned more wealth than the bottom 50% of a planet of 7 billion. Conflate that with the 47% claim and for some you have the makings of Feudalism 2.0.
While that's possible, the economists who have looked into the issue present a more optimistic scenario. Osborne, for instance, predicts that many of the 47% of jobs marked for extinction will be replaced by other, new jobs.
"Our society has undergone enormous technological change over the last few hundred years, without corresponding increases in unemployment," he says. "For example, over the 20th century, the fraction of workers in agriculture in the U.S. went from around 40% to less than 2%, but, nonetheless, those workers found new jobs in industries not even imagined at the beginning of the century."
Brynjolfsson is similarly upbeat, noting that too many subscribe to the false narrative that technology controls us. Actually, it's the other way around, he says. "I could see a world where we have a lot more leisure," he says. "In 2034, extreme poverty could be eliminated globally. It's within our reach." However, "I could also see a world where there's more inequality. We need to change the conversation to see what kind of outcome we want. We have a lot more power than we think."