Marissa Mayer studied AI in the ’90s before joining Google

When I arrived at Stanford, I was pre-med, and I was certain I was going to be a pediatric neurosurgeon. I realized over the course of my freshman year that I was less interested in cutting up brains, and I was much more interested in how they worked. I found a major, Symbolic Systems, which is a psychology, philosophy, linguistics, and computer science that basically studies how do we learn cognitive psychology, and I was hooked. There was a specialty offered in my major of either neuroscience or artificial intelligence. I got really excited about what could come of artificial intelligence. Ironically, I was reflecting back, my honors thesis was a voice-assisted travel planner. The moment is really here where we're seeing models that are sophisticated enough that they can really mirror human intelligence. For a long time in computer science and technology with AI, we were focused on the learning and the reasoning. When I think back on Google Translate, when I think back on Google Spellcheck, when I think back on the early days of Google Assistant, Those were the things that we were focused on. And it really took the spark of applying transformer models and allowing them to express themselves, to talk back to you. That was a surprise to me that that was really the catalyst for it to become mainstream.