Introduction and path into Anthropic

0:00 With experience spanning computer science, biotech, and machine learning, Hansel  brings a unique perspective to building safe and reliable AI. Join me as we dive into Hansel's  journey and explore anthropics approach to shaping the future of artificial intelligence. Welcome,  Hansel, to the Getting to Alpha podcast. Thank you for having me. It's a pleasure to be here. So  excited to have you here. Let's start by winding it back. What was your educational background  that led you here right to anthropic in the eye of the hurricane?

1:02 I started studying computer  science at Stanford in undergrad. Originally, I'd actually been pursuing theoretical physics  and then chemical engineering and then finally took a programming class and absolutely loved  it and sort of naturally along the way found that of all the different branches I could  pursue in computer science, AI at the time seemed the most fascinating and really the  most fun. And this was back in 2015, I think. And so from there I continued to study with a  specialization in AI which at the time didn't really mean neural networks so much.

1:33 It had  started to move in that direction but it wasn't obvious. From there after school I joined  Nvidia where I worked for about 2 and 1/2 years primarily using machine learning and AI to try  to improve their chip design. And then at that point I actually jumped into biotech as a result  of I think really the pandemic and wanting to work directly with companies that were working  on responses to CO that lasted another 3 years at which point I I took a look around and saw that  the entire landscape of AI had changed completely.

2:01 left biotech and started looking around at labs  and companies that I wanted to work at. And I think Anthropic really stood out at the time as  heavily differentiated in the type of research they wanted to do and why. And they were pushing  very hard in the direction of AI interpretability, AI safety. In some senses, even like a level of  AI philosophy and trying to understand what kind of impacts we might expect from this technology  beyond just the bottom line profits.

2:31 So I can do hyperdimensional math and I can do game theory in  the same place and I can take in like very complex like high performance computing and distributed  systems and all these things mesh together in ways that you absolutely would not expect to point that  conversations with colleagues about philosophy are actually fairly common in the hallways because the  intersection of all the different sort of relevant ideas and concepts here is very wide. So that was  extremely appealing.

3:00 I ended up interviewing there and have been very excited to be working there  ever since. So tell us about the different roles you've played at Anthropic and also what you're  doing in your current role. Absolutely. So when I first joined Anthropic, I actually joined as  part of the inference team and the core sort