I’ve been building AI agents but primarily had a focus on a gameplay agent for Caves of Qud. It’s led me to question the true nature of memory and consciousness which are lightly explored in some of the blog posts; and expanded upon in a series of essays that I am still working through. During that exploration I have tested some theories in different cognitive architectures and will look to show that experimentation too, even where I managed to replicate the convergence rate of MAMBA with the long term stability of Transformers at the cost of 10x the training speed.

Langs

Most of the projects and work are in:

• Python
• Rust
• C++
• Typescript

I am open to whatever the particular job requires.