Bringing Sloan Savvy to an AI Chatbot Startup

How do you go from a humanities degree at Har­vard to a video-game com­pany in London to an AI/machine-learning startup in Boston? Stop at MIT Sloan along the way.

“I had this stereotype of video-game designers as boys in basements,” says Catharina Lavers Mallet MBA ’09, who joined Playfish, a social games startup in London, after her time at MIT. She quickly discovered that the fast-moving industry’s blend of creativity, innovation, and data was infectious. She wound up managing production of games with multimillion-dollar bud­gets, including the Sims Social. She joined King Digi­tal (then King.com) in early 2012 to build and expand its UK game studios dur­ing the company’s explosive growth, driven by the suc­cess of Candy Crush Saga.

The greatest takeaway was self-confidence.

Mallet’s degrees were both useful. “Harvard was a great experience in teach­ing me how to think, and Sloan was a powerful con­trast because it’s so action-oriented,” she says. “You have to think, but you also have to show what you can do. The greatest takeaway was self-confidence—I had never come across certain problems before, but I knew I could figure out how to get it done.”

In 2016, after sev­eral years in London and a period of travel, Mallet decided to relocate to Bos­ton with her family. Through networking and with help from the Martin Trust Cen­ter for MIT Entrepreneur­ship at Sloan, she landed as COO at Talla, founded in 2015 to develop AI-based chatbots that act as specialized assistants for knowledge workers. The company eventually hopes to become a digital plat­form that can take on much of the repetitive work peo­ple do, freeing them up for more valuable, creative, and strategic activities.

Is Mallet concerned about the potential for AI and automation to destroy jobs?

Fears about AI-based technology were exacer­bated by reports that Rus­sian Twitter bots created news to sway the 2016 US presidential election. To address such fears, Talla is developing blockchain-based software to verify the authenticity and credibility of bots for business-to-business AI applications.

Is Mallet concerned about the potential for AI and automation to destroy jobs? “We’re many years away from workers’ facing replacement by bots. Today, bots can provide very real incremental assistance to people,” she says. “Longer term, we have to focus on making sure people have flexibility in their skill sets and that we’re preparing students for these rapidly evolving technologies.”

First published in MIT Technology Review, March/April 2018.