Hello, Alistair here. I recently chatted with the incoming head of Google's advertising business, Prabhakar Raghavan. On Friday, he takes over from Sridhar Ramaswamy to run a business that generates more than $100 billion in revenue a year, making the 58 year-old the most important person in advertising. Here are three main takeaways from our conversation:
1. Raghavan is no Don Draper. Raghavan, who has a Ph.D. in electrical engineering and computer science, is an academic at heart. He will apply his deep technical knowledge of search, web link analysis, artificial intelligence—and a field of study known as market design—to invent new ways of running ads online.
"I don't claim to be particularly creative," he said. "My roots are in hard computer science, including search and advertising, stuff I did way back at Yahoo and other places including back in the time of Larry (Page) and Sergey (Brin). There are interesting challenges in market design. I think of this area as full of opportunity, rather than `OK here's a feature that's going to make income.'"
2. Raghavan and Google co-founders Larry Page and Sergey Brin go way back. Raghavan was a consulting professor at Stanford University when he met Page and Brin in 1996. They were all studying how to retrieve information from the nascent world wide web at the time. So he's likely to stick closely to Google's original mission of organizing the world's information, a task he says it still far from complete.
"I recall those days quite well because these were deep technical discussions and problems that I had my team working on and they had their team working on. We would talk about search as a challenge for the ages," Raghavan said. "What struck me was Larry's conviction that search would remain an unsolved problem for a long time. He has been proved right."
3. Raghavan will be a heavy user of AI in his new job, as he was in his old one.
"I did not expect in my lifetime that computers would beat humans at Go," he said. "That happened about 50 years early. There's a confluence of three things: massive cloud infrastructure at scale; large amounts of data in the cloud; and recent algorithmic improvements, deep neural nets especially, that have been astounding, almost inexplicably so. We can now re-imagine what is solvable by machines."
It's worth noting that Raghavan helped create Google's AI-powered features like Smart Reply, which suggests appropriate responses to emails for Gmail users. "Today, over 10% of English Gmail replies are machine-written but human-accepted," he said.—Alistair Barr