When humans first discovered fire, the one who knew the secret of creating it was the most powerful person in the tribe.
With a zippo in every pocket, new differentiators have to be found.
In the past, when a job was automated, it was usually around physical tasks: harvesting wheat, operating an elevator, moving boxes.
Some kind of mechanical device got built – a combine, elevator buttons, a forklift – and that job shifted.
Now, as computers are getting better and better at seeing and following patterns, and as more and more of our world is networked and API-friendly, we’re seeing more and more mental work outsourced to algorithms.
Is everything going to be automated? Why is this happening now? What’s the future going to look like for humans? What will be next?
There are a couple roadblocks to automation… Computers are dumb, and they don’t act like humans.
Computers are still pretty limited in what they ‘understand,’ with most of the complicated and unstructured situations being the most difficult to grok.
This leaves the repetitive jobs for the algorithms: is there an x in this picture? does this text match something we’re looking for? what are the trends in this batch of data? which of these things is an anomaly?
Any task that can be broken down into a formula is ripe for automation.
All tasks which can be done over and over and over again – they “scale” well. (For software.)
Every job is full of these kinds of tasks, even as the details change.
Computers are also not humans, so most of the jobs that they’re really good at involve interfacing with other computers or with a trainer or manager to help guide them. With an in-the-know human in the loop, the architects don’t have to worry about making the algorithm look friendly, or making it operate 100% perfectly, or making it totally resilient to all sabotage. The human can take over for weird blips in the line, and the computer can handle all the routine stuff.
On the count of computers being dumb… Why are they getting smarter? How come it feels like they’re doing more to automate tasks than in years past?
Probably a mix of things: better hardware, better networking, and better algorithms. Our computers are faster, more connected, and we’re able to train them better than before. Also, as an engineering culture with ‘open source’ as a major tenet, many of our major software successes can be duplicated around the world within a few short hours.
The amount of automation that it takes to reach a sort-of technological equilibrium varies.
Looking back at the old jobs for a moment…
The elevator operator’s job split to a few people: the button maker, the elevator repairer, and the ux designer. The task of “which floor?” gets outsourced, but it requires a touch more work to make the job feel complete. Once that work is complete, the only ongoing task is making sure the whole machine keeps running, and maybe designing and building new machines every dozen years.
The wheat harvester got outsourced by a machine which wiped out dozens and dozens of jobs in favor of a few: the combine manufacturer and the combine operator. The combine operator, an entirely new position, is really just steering a few tons of mechanics, keeping an eye out for anomalies, and keeping the rubber side on the ground. In newer versions of the combine, there isn’t even a pilot in the seat: herds of combine drones can be controlled and managed back in the farm house, leaving the steering and most of the routine problem solving to the combine itself.
The box mover and their fork lift is in a similar boat: lots and lots of lifting and moving goes on in an automated Amazon warehouse, but most of it is by robots. A few humans at the head of the warehouse pull products off a shelf and place it into a box, but the shelves and the boxes and the items all come to the human, automatically. As automation gets better, Amazon will eventually automate that task, as well, leaving… who in the warehouse? Drivers, managers, and maintenance engineers, mostly.
What ends up happening when jobs are automated?
We get robots doing old jobs, and humans doing new jobs: manufacturing the robots, managing the robots, and maintaining the robots.
Along with those obvious new jobs, though, we sometimes get lots and lots of opportunity. Wheat is plentiful, Amazon is cheap, and anybody can operate an elevator.
Who benefits? Who is selling the shovels to the gold miners?
How might we intelligently grow the new robo-labor to benefit the whole of society in greater ways?
Nichols believes that workers will ultimately prefer an environment where they feel part of the larger business and can acquire new skills. “Look at any company structure and there’s a ladder to climb where people get paid more as they move up. There’s a similar structure to be built on top of the crowd,” he says. “I’d love to see the day where someone can have a career in crowdworking: to work whenever they want, wherever they want, and get paid more for their experience.”
Google’s Paritosh is a big fan of this human-centered approach to crowdworking. “You use the technology and the platform to connect and communicate, but you’re essentially building a human organization,” he says.
Ryo Suzuki of the University of Colorado Boulder tried to do just that. He developed a ‘micro-internship’ platform called Atelier that paid experienced programmers to mentor newcomers on a week-long e-commerce coding project. In its first test, heavier use of Atelier was associated with higher quality work, and the interns reported learning new skills.