Jason Lo

Absolutely not necessary. More like a ‘because I can and no one can stop me’ moment. github.com/JasonLo/claude…

The evolution of programming: Problem → Code → Solution 😥 Problem → Code (AI assisted) → Solution 😄 Problem → Rigid parts → Code ┐ ├→ Solution 🤯 Problem → Abstract parts → AI ┘

We’ve had a very positive reaction to RABBIT from our community. Shoutout to our lead developer @JasonLoCM

A small personal project inspired by @karpathy 's autoresearch: jasonlo.dev/projects/skill… Instead of randomly adding Claude skills, the system continuously discovers, evaluates, and improves skills tailored to user needs.

Our website devoted to collecting resources and papers on simulation-based inference got an upgrade. @JasonLoCM is using ChatGPT to categorize papers (not everything is on the arXiv). @datascience_uw simulation-based-inference.org/papers/

Jason Lo joined the team at the Data Science Institute this winter. He specializes in crafting custom deep neural networks that simulate children's reading acquisition. Read his story in Faces of Data Science: #faces" target="_blank" rel="nofollow noopener">datascience.wisc.edu/#faces @JasonLoCM