I write on topics such as forecasting, effective altruism, tech, finance, and more broadly “understanding and optimizing the world.”
Some suggestions for articles to check out first:
Comparing election forecast accuracy - What can we learn from scoring different election forecasts?
My collection of Covid articles - The best Covid protection measures you should take
I’ve been writing research and analysis on these topics for myself and a few friends for a little while, and I figured other people might find it interesting and useful too - so here you go!
The inspiration for the name “First Sigma” is in contrast to the two to five sigma confidence (p-values) often used in the sciences. Many of the most difficult and important topics in the world don't have enough research and data to put anything down with high confidence, but are still very much worth our best effort. It’s also inspired by my practice of making specific predictions, with probabilities and confidence intervals, about these tricky topics. I use my generalist forecasting expertise to integrate together hard-to-interpret information from subject-matter experts, models, news, etc.
Why should you believe my predictions? By putting specific numbers on a lot of predictions and seeing how accurate they ended up being, I can compare my track record to other experienced forecasters, and see evidence that my predictions are better than most:
I’m a high ranked forecaster on Metaculus and Manifold (at the time of this writing, I’m ranked #1 on Manifold).
I’ve participated in several forecasting efforts across a range of topics, such as the AI Progress Forecasting Tournament where I won #1 for on making a broad range of forecasts on AI performance, research agendas, and investment over a 2-year timeframe; as well as Covid (Metaculus project and #1 on Manifold) and politics (#1 on Manifold on 2022 US elections).
