Neel Nanda did not expect to be leading a team at Google DeepMind at the age of 26.
A self-described perfectionist, the lead for Google DeepMind’s mechanistic interpretability team said he often hesitates before starting projects, afraid they might flop.
“I often don’t want to do things. I’m like, ‘This seems risky. This could go wrong,'” Nanda said on an episode of “The 80,000 Hours Podcast” published Tuesday.
To break that perfectionist streak, the senior research scientist said he challenged himself to write one blog post a day for a month.
Nanda said this routine helped him build visibility in the AI community, produce work that shaped the field of mechanistic interpretability, and even meet his partner.
“One of the most important lessons I’ve learned is that you can just do things,” he said. He sees it as part of “maximizing your luck surface area,” saying yes more often and giving luck more chances to show up.
Sometimes that meant doing “things that are kind of a bit weird or unusual,” like posting a three-hour unedited YouTube video of himself reading through a dense AI paper, he said.
Related stories
Business Insider tells the innovative stories you want to know
Business Insider tells the innovative stories you want to know
It ended up drawing more than 30,000 views.
That same mindset led him into leadership. When Nanda joined DeepMind in 2023, he expected to be an individual researcher, he said. But a few months in, the team’s leader stepped down — and Nanda agreed to take over.
“I did not know if I was going to be good at this,” he said.
“To me, this is both an example of the importance of having luck surface area — being in a situation where opportunities like that can arise — but also that you should just say yes to things, even if they seem kind of scary,” he added.
Nanda did not respond to a request for comment from Business Insider.
