Hey! Nice to meet you!

Hi, I’m Cody! Human first, philosopher and computer scientist second.

I study Computer Science at the University of Texas at Austin, and currently do Value Alignment Research with Brad Knox and Control Research with Buck Shlegeris.

In the past, I’ve worked on AI Mechanistic Interpretability research under Neel Nanda, as well as in Cybersecurity and Game Development. Nowadays, most of my time is spent thinking about the future and how we can develop safe Artificial General Intelligence. Besides all that, I’m working on being a better climber, writer, thinker, and friend.

What am I currently doing? May 2024 Edition

  • I am taking a break from Mechanistic Interpretability research and try new research out! I've been working on RL Value Alignment with Brad Knox, and will be in the Bay Area starting June at MATS again doing Control Research with Buck Shlegeris
  • I’m still reading The Path To Power. It's a long book and I move slowly lol, but I'm moving faster now. It's a fascinating book. I'll have a blog post on it when I'm done reading.
  • I'm learning old, fundamental Machine Learning topics to try to develop better intuitions about learning more broadly (with this book, Ilya 30u30, and other sources)
  • I'm overcoming my climbing plateau! Stuck trying to consistently get V4s, but I WILL NOT PLATEAU PERMANENTLY
  • My Self-Repair paper (link) got accepted into ICML! Debating on if I will go or not, but I will also be reviewing for the ICML Mechanistic Interpretability workshop
  • I'm off of school now and am thinking of taking a gap semester in the fall to do something cool. Message me if you have any ideas :)
  • I'm experimenting with friends on using 'anti-charities' as a means of ensuring productivity, and it is working amazingly