Catherine Olsson ’08 

Senior Program Associate, Open Philanthropy | Potential Risks from Advanced Artificial Intelligence

July 2020

by Jim Collins, photo by Jason Arthurs

The Spring/Summer 2020 issue of Lakeside magazine featured "The Futurists," a gallery of forward-looking alums who, unconstrained by current assumptions, offer a view from outside the box.

BACKGROUND // Research Software Engineer, previously worked at Google Brain and OpenAI, degrees in computer science, brain & cognitive science, and neuroscience.

HER QUESTION // If you imagine a future in which systems are powerful enough that companies can use virtual staff to do knowledge work, these systems might be put in charge of important business or policy processes at large scale. If any of those systems put the full weight of their intelligence behind a course of action that’s a big mistake, will the resulting mess be beyond human beings’ ability to halt the cascading disruption?

THE WORK // Grantmaker at Open Philanthropy. Funds technical research that could make us better prepared for the possible AI future.

THE FUTURE // New paradigms in machine learning will emerge, and we will need new research leading to insights for how we might predict, understand, and supervise more-advanced systems.

IN HER WORDS // “It’s important to recognize that technological accident risks threaten all of us collectively. A parochial, nationalistic, or adversarial lens is not the way forward. Global collaboration will be crucial.”