Nathan Douglas
Hi, I'm Nate, or Nathan, Nug Doug, Nug Diddy, whatever.
I'm really interested in myself, so I've decided to create a website devoted to me. I'll read or program or whatever, and then I'll write about it on here. Look on my works, ye mighty, and close the tab.
I'm a very curious guy with strong interests in complexity science, the social sciences, organizational dynamics, system design, computability, artificial life, and... I guess the list goes on.
I'm currently working as a Senior DevOps Engineer (though in practice more of a combination software and infra engineer), but I'm a generalist and comfortable in a lot of different scenarios, areas, and niches.
The background stuff is an ant colony simulation. When I gaze into the abyss, I like to see something soft and subtle and alive.
Although I should know better, you can also find me on social media:
If you're interested in working with me, here's some professional information:
Recent Posts
Recent Reading
- Human Compatible, Stuart Russell: A comprehensive exploration of the challenges and opportunities in developing artificial intelligence that aligns with human values and interests.
- L.A. Confidential, James Ellroy: A crime thriller set in 1950s Los Angeles, following the story of three compromised detectives.
- Neuromancer, William Gibson: A dystopian science fiction novel set in a future where cyberspace is a real place, following the story of a hacker named Case.
- Piranesi, Susanna Clarke: A man living in a labyrinthine house filled with statues and endless corridors, exploring the mysteries of his environment and the nature of reality itself.
- Procedural Generation in Game Design, Tanya X. Short and Tarn Adams: A comprehensive guide to the principles and techniques of procedural generation in video game design, offering insights into creating dynamic and engaging game worlds.
Recent Ideas
- Event sourcing/event log-based infrastructure management; replace the declarative state structure of IaC with a declare-suggested-changes structure that is merged into an event log-based immutable record.
- Can we do machine learning with a neural network that has multiple depths? e.g. a 30-layer network that also has pathways that will skip multiple layers? Thinking about this as an analogue of focused vs. diffuse thinking.