Meet the Humans Preparing AI for the Robot Job Revolution

Friday, 28 April 2017 - 2:35PM
Technology
Robotics
Artificial Intelligence
Friday, 28 April 2017 - 2:35PM
Meet the Humans Preparing AI for the Robot Job Revolution
Image credit: Fonytas
If science fiction has proven anything, it's that humanity is never prepared for the robot uprising. But leaving aside the prospect of a violent coup by Terminators, the biggest danger may be the obsolescence of human workers as artificial intelligence continues to grow in complexity and popularity. Machine learning protocols can be used for anything from personal assistants to customer service roles, and just as manufacturing replaced human workers on assembly lines, it won't be long before humans are phased out of plenty of "critical thinking" jobs in favor of the cold, hard logic of robots. This time, however, humans are voluntarily teaching robots how to do their jobs.

A recent article from the New York Times profiles five workers in various fields who've become teachers to artificial assistants. Coming from an impressive range of backgrounds (including a former army captain and a playwright), these AI tutors are responsible for breathing life and nuance into the workings of their digital successors, making their services all the more useful. For example, there's an AI chatbot that's designed to make lightning-fast travel recommendations based on customer profiles, a bot that translates legalese into something a little more readable, and a self-driving car that's learning to spot unusual hazards and react accordingly.

What's interesting is that these bots are capable of either perfectly mimicking human interactions or even surpassing the skills of their flesh-and-blood creators. Interaction designer Diane Kim has expressed excitement at her personal assistant's ability to pass an informal Turing test with clients: "People ask them out on date. They receive thank-you emails from happy customers even though, as robots, they don't need gratitude."

And they're learning fast. Rachel Neasom trains a travel recommendation bot named Harrison, and has already seen it succeeding in unexpected ways:

Opening quote
"At first, Harrison would recommend hotels based on obvious customer preferences, like brands associated with loyalty programs. But then it started to find preferences that even the customers didn't realize that they had. Some people for example, preferred a hotel on the corner of a street versus midblock."
Closing quote

It's this kind of pattern recognition that makes machine learning such a valuable asset: AI programs can spot trends that the human mind simply can't keep track of. But if the computers are evolving, what are humans to do about their own jobs? The solution, it turns out, is to evolve as well. Once upon a time, people like Rachel Neasom would have been effective travel agents themselves. Now, it's more important to be able to utilize AI tools to work faster and smarter in the kinds of jobs that robots won't be able to learn just yet.

Another article from the New York Times, by Cixin Liu, points to the importance of creative work in the future—when AI programs can tackle accountancy, banking, and scientific research, it'll be the inventive, creative humans who thrive most. There's particular potential around human interactions: robots can be programmed to react to people, but they struggle to spot nuances in the way we interact—after all, computers can be programmed to understand logical processes, but there's not always a lot of logic behind human behavior.

Ultimately, humans are best at understanding humans, and that's a skill that can't be replicated by AI just yet. So while a lot of jobs will be going to the machines in the coming years, the future's bright, so long as you're a people person.
Science
Science News
Technology
Robotics
Artificial Intelligence
No