Will a robot take your job?

There is a vigorous debate about the effects of automation on jobs. Everyone agrees that some jobs will be lost to automation and, in turn, some jobs will be created by it. The pivotal question is how all of that nets out.

Often lost in the abstract debate is the question of exactly which jobs are likely to be automated. I have created a test to try to capture just that.
The idea is simple: Some things are quite easy for computers and robots to do, and other things are quite hard. Jobs in the “safe” category have lots of things about them that are hard for machines to do.
The good news is that it doesn’t take very many hard things to make a job, practically speaking, impervious to automation, at least in this century. While jobs like “hostage negotiator” are clearly better done by people than machines, even jobs that look like good candidates for automation have difficulties. In theory, a robot should be able to clean the windows on my home, in practice this isn’t likely to happen for quite a long time.
The test is ten questions, and each one can be scored from 0 to 10. For each one, I give examples of some jobs at 0, 5, and 10. My examples are meant to show each extreme, and a midpoint. You should not just score with those three points. Use 7’s and 2’s and 9’s.
When you are done, the total is tallied. The closer it is to zero, the less likely you are to get a surprise announcement from the boss one day. The closer you get to 100, well, if you start to feel something breathing down your back, then that may be the cooling fan in the robot who is about to take your job.
The goal is not to find a job near a zero. Anything below a 70 is probably safe long enough for you to have a long illustrious career. There are obvious “100” jobs. The person who takes your order at a fast food restaurant is probably pretty close.
Take the test here. We plan to calibrate and refine it, then publish a research report about the results. If you would like to be kept in the loop about that, be sure to add your email address.

Report: Uber hired 50 scientists from Carnegie Mellon to build self driving cars

We knew this day would come, but we didn’t know it would be this soon.

According to TechCrunch, Uber is building a research facility in Pittsburgh to invent its own self-driving cars. Not content to rely on Google, the on-demand ride company has reportedly recruited researchers from Carnegie Mellon University’s Robotics Institute to build the product.

It has staffed up with 50 senior scientists who will work on the software technology and the vehicles themselves. The Robotics Institute has been “cleaned out” with the flood of high profile departures, the TechCrunch report said.

After this story published, Uber released a blog post confirming the news. It called the center a “partnership” between Carnegie Mellon and Uber. The building will be called the Uber Advanced Technologies Center. As part of its development, Uber will fund faculty chairs and graduate fellowships at Carnegie Mellon. And in the blog post announcing the news, Uber included supportive quotes from Carnegie Mellon’s dean of the computer science department and the Pittsburgh mayor.

Uber CEO Travis Kalanick has said for a long time that the company intended to eventually shift to self-driving vehicles. That would cut a huge chunk of its revenue cost — drivers take 80 percent of every transaction. “When there’s no other dude in the car, the cost of taking an Uber anywhere becomes cheaper than owning a vehicle,” he explained at the Code conference in May.

Many have used Kalanick’s remarks as proof the company doesn’t care about its drivers’ well-being, since it hopes to eventually eliminate the need for them. On the flip side, the convenience, safety, and efficiency for passengers would be substantial.

A shift to self-driving cars would fundamentally change the nature of Uber’s business by putting the company in charge of the vehicle fleets it deploys. Until now, it has acted as a transportation platform, connecting willing riders to willing drivers, but not owning the hardware of the operation itself.

According to the report, Uber has already started to build work stations for the scientists, although there’s no word on the project’s timeline for completion.

Move over Emeril: Robot learns how to prep food from YouTube

These days, you can learn just about anything from YouTube videos — from how to tie a knot to the best way to open a wine bottle with a shoe. And it isn’t just humans who are benefitting. University of Maryland researchers have programmed a robot to learn basic cooking skills from YouTube videos, a feat that could eventually be expanded into other skills like equipment repair.

The researchers worked with Baxter, a robot built by Rethink Robotics. Baxter is popular for its safety around humans and ease of use; it can be programmed quickly just by moving its arms, and it is smart enough to adapt to a changing work area.

But Baxter’s default software isn’t smart enough to understand a video, let alone recognize the correct measuring cup or ingredient. With the Maryland team’s help, Baxter was able to watch YouTube videos and learn what types of objects to recognize, catalog directions by picking out action verbs and observe which type of grasp would be most effective for holding each tool. Baxter then repeated the steps in the video without any input from its human operators.

University of Maryland computer scientist Yiannis Aloimonos (center) observes as Baxter measures ingredients.

University of Maryland computer scientist Yiannis Aloimonos (center) observes as Baxter measures ingredients.

“This system allows robots to continuously build on previous learning—such as types of objects and grasps associated with them—which could have a huge impact on teaching and training,” DARPA program manager Reza Ghanadan said in a release. “Instead of the long and expensive process of programming code to teach robots to do tasks, this research opens the potential for robots to learn much faster, at much lower cost and, to the extent they are authorized to do so, share that knowledge with other robots.”

DARPA, which funded the project, has other applications in mind like military repairs and logistics. Now that Baxter is cooking, there is no word on when the robot will learn to do the dishes.

DARPA’s updated ATLAS robot runs on battery, is even more badass

Those disappointed by the glacial pace of the competitors in last year’s DARPA Robotics Challenge trials will appreciate the newly updated ATLAS robot DARPA debuted today. ATLAS is the generic robot body used by teams that concentrate on software, and therefore has a very strong presence in the competition.

The most obvious change in the ATLAS robot, which was built and now updated by Google-owned Boston Dynamics (which pulled its own robot from the competition), is it now has a shiny white chest plate that makes it look like a cross between a Stormtrooper and the Terminator. But the biggest upgrade is its power now comes from batteries instead of a tethered cable, which will allow it to move more freely in the competition.


According to DARPA:

Atlas will now carry an onboard 3.7-kilowatt-hour lithium-ion battery pack, with the potential for one hour of “mixed mission” operation that includes walking, standing, use of tools, and other movements. This will drive a new variable-pressure pump that allows for more efficient operation.

Free movement is important because the DARPA Robotics Challenge is actually a demonstration of the technology that will someday evolve into search and rescue robots. When a robot enters a burning building, it can’t be moving at two miles per hour trailing a power cord. It needs to be capable of quick autonomous action even in the face of rubble and the collapse of communication networks.

Overall, only ATLAS’s lower legs and feet remain the same. It has more flexibility in its wrists and can now see its hands at work. It’s also stronger and quieter.

The Robotics Challenge finals will take place in June in Pomona, Calif., where the winning teams will split $3.5 million in prize money. At least 20 teams are expected to compete, and the competitors will be different than last year’s mix.

The newest version of the ATLAS robot runs on battery, instead of a tethered power source.

The newest version of the ATLAS robot runs on battery, instead of a tethered power source.

Forgoing the power cord brings ATLAS in line with changes DARPA is requiring for the competition’s finals. The DARPA announcement states:

  • Robots will have to operate completely without wires — they may not be connected to power cords, fall arrestors, or wired communications tethers. Teams will have to communicate with their robots over a secure wireless network.
  • Teams are not allowed any physical intervention with their robot after it begins a run. If a robot falls or gets stuck, it will have to recover and continue with the tasks without any hands-on assistance. If a robot cannot sustain and recover from a fall, its run will end.
  • DARPA will intentionally degrade communications between the robots and human operators working at a distance. The idea is to replicate the conditions these robots would face going into a disaster zone. Spotty communication will force the robots to make some progress on their own during communications blackouts.

Hopefully the change in rules will push teams to be more daring, which will lead to both greater successes and failures.

Rethink Robotics nabs $26.6M to keep building easy-to-train bots

Rethink Robotics, builder of the beloved Baxter robot, announced a $26.6 million Series D round today, bumping up its total funding to more than $100 million.

GE Ventures led the round. Goldman Sachs, Bezos Expeditions, CRV, Highland Capital Partners, Sigma Partners, Draper Fisher Jurvetson and Two Sigma Ventures also participated. Rethink Robotics didn’t go into details about the funds’ use, but said the money will go toward international expansion and research.

Baxter is a human-sized robot that can be quickly trained to complete a new task by dragging its arms from place to place and telling it to repeat. It is safe to operate around humans and is popular in both industrial settings and laboratories. It sells for $25,000 — not the cheapest these days, but far below the cost of traditional industrial robots.


Rethink Robotics announced a major software update in November that uses markers, similar to barcodes, to help Baxter recognize and interact with its environment. The robot uses the markers to keep track of items even after they are bumped or moved and alternate between tasks without having to be retrained.

Need more battling robots in your life? There may be a solution

Two robots battle at the 2013 RoboGames. Photo by Signe Brewster.

Two robots battle at the 2013 RoboGames. Photo by Signe Brewster.

First BattleBots fell. Then Robot Wars. For a whole decade now, the world has been woefully short on a fighting robots show. A new Kickstarter campaign intends to change that by creating a video series based off the wheeled robot battles at RoboGames, an annual event that features many different types of bots. If it succeeds, sparks, fire and a whole lot of epic robots will head to screens in July 2015.