A look at five people training the robots that might replace them
“Meet the people who train the robots (to do their own jobs)” is the catchy title of an article in The New York Times, by Daisuke Wakabayashi. Two of the five people he profiles—a travel agent and customer-service representative—seem likely to face the most competition from the robots they are training. The chief executive profiled seems unlikely to be replaced—he just wants robots to help cut back on the hundreds of dollars per hour he has paid lawyers to look for capitalization errors in contracts. Finally, a software engineer and an “interaction designer” (which Wakabayashi describes as “a scriptwriter, of sorts) seem engaged in long-term efforts to improve the robots they work with.
The travel agent, Rachel Neasham, works with a robot named Harrison at Lola. She describes being in a race with Harrison, who can quickly book a hotel room but can’t advise a family vacationing at Disney World on the best strategy for getting an unobstructed photo of the children in front of Cinderella Castle—at least not yet. Wakabayashi quotes her as saying, “It made me feel competitive, that I need to keep up and stay ahead of the AI.”
The customer representative, Sarah Seiwert, works at Magoosh, an online test-prep company, answering student emails. The company uses software from DigitalGenius to speed the process of finding the appropriate preapproved response from a database of thousands.
The software hasn’t replaced human customer representatives but has cut the backlog of customer requests by half. “I am not convinced that artificial intelligence is going to replace us,” Wakabayashi quotes her as saying. “You can’t program intuition, a gut instinct.”
Diane Kim, the interaction designer, has a job at x.ai that Wakabayashi describes as “part playwright, part programmer, and part linguist.” The company offers human assistants—Andrew and Amy Ingram, both of whose initials are AI—that schedule meetings using as few email messages as possible yet without coming across as condescending. An offer to put a prospective attendee on the boss’s calendar, for example, came across as cold. In addition, the AIs need to deal with ambiguities—distinguishing between a request to set up a meeting on “Wednesday” vs. on “a Wednesday,” for example.
x.ai doesn’t pretend its AIs are human, Wakabayashi writes, adding, “But Ms. Kim still gets satisfaction when people don’t realize that the assistants are robots. People ask them out on dates. They receive thankyou emails from happy customers even though, as robots, they don’t need gratitude.”
Dan Rubins is chief executive of Legal Robot, which uses AI to review legal documents, flagging anomalies. Wakabayashi quotes Rubins (who is not a lawyer) as saying, “Lawyers have had 400 years to innovate and change the profession, and they haven’t done it. It’s time for some outside help.”
However, Rubins added, “I really don’t think we’re going to get rid of lawyers. Unfortunately, we still need them.”
Finally, Wakabayashi profiles Aleksandra Faust, a software engineer at Waymo. Jobs like hers would seem secure as engineers working on autonomous vehicles will continue to face novel situations involving safety and comfort. Wakabayashi quotes her as saying, “One thing we’ve learned about human driving is that it’s very, very complex.”
