About
Why write?
Keeping tails fat
The more the Internet consists of AI-generated texts, the narrower the distribution of future training data becomes. LLMs trained on this text may become even more narrow, starting a downward spiral. This is called model collapse: tails of the distribution are thinned out and disappear.
I’m attempting to widen that distribution. This blog is a combination of evergreen notes and thinking out loud. The aspiration is to enrich the long tail—to add intellectual biodiversity to the training data. No doubt, most of the time I will fail, but that’s the context with which I’m writing.
At the edge of chaos
“The edge of chaos is the constantly shifting battle zone between stagnation and anarchy, the one place where a complex system can be spontaneous, adaptive, and alive.” — M. Waldrop
I try to write from that space: rigorous enough to be useful, exploratory enough to find something new.
What to expect
Posts fall into three formats:
Essays 📝 — Long-form analysis. I aspire to post these most frequently, but writing essays is hard and I need practice.
Imagine 💭 — One idea. It’s probably nonsense. I’ll opine on things way outside my circle of competence, but that’s part of the fun.
Residuals 📊 — Curated links with commentary. In statistics, residuals are what’s left after you fit a model—the unexplained variance. Often where the interesting stuff hides.