Neel Nanda is just 26 years old, but he's already leading Google DeepMind's mechanistic interpretability team. He never expected to get such a big role this early in his career. Nanda shared his story on The 80,000 Hours Podcast, on Tuesday, September 16, 2025. He said his secret was learning to say yes to opportunities, even when they seemed risky.

Fighting perfectionism

Nanda admits he's a perfectionist who used to hesitate before starting projects. "I often don't want to do things. I'm like, 'This seems risky. This could go wrong,'" he said on the podcast.

To change this habit, he challenged himself to write one blog post every day for a month. This daily writing made him more visible in the AI community and helped him develop ideas for his research. It even led him to meet his partner.

"One of the most important lessons I've learned is that you can just do things," he said.

Creating opportunities

Nanda calls his approach "maximising your luck surface area." This means saying yes more often to give chance more room to work.

He once uploaded a three-hour unedited YouTube video of himself reading an AI paper. It got over 30,000 views. "People were into it," he said.

Unexpected leadership

According to a Hindustan Times report, when Nanda joined DeepMind in 2023, he planned to stay as a regular researcher. But when his team's leader left after just a few months, he was asked to take over.

"I did not know if I was going to be good at this," he admitted. But he said yes anyway.

Now he leads a team focused on AI safety and has published many papers. He has also mentored over 50 junior researchers, with seven now working at major AI companies.

Nanda says people overestimate risks and underestimate their ability to recover from problems. His advice is simple: "You can just do things." He believes today's young researchers have an advantage with AI tools that can help them learn faster than ever before.