Loving Claude Fable 5's strong safety measures protecting me from exploring extensions of ideas related to epsilon-machines+computational mechanics and deep learning.
I liked ChatGPT Pulse but cancelled my Pro subscription a few months ago.
So I vibe-engineered a pi + Ollama Cloud pipeline for creating Pulse-like reports from chatbot conversations, coding sessions, and Anki cards.
Works surprisingly well for a 2-weekend project.
Link below.
It is kind of crazy I can just say "Hey Claude Opus 4.7: I want an open-source version of ChatGPT Pulse that processes my chatbot conversations, agentic coding sessions, and Anki notes through Pi using Ollama Cloud" and it gives me workable first-draft in less than an hour.
I made a thing. I've been using Claude Code, Codex CLI, and Cursor and wanted one place to browse and search all my coding sessions across coding clients.
Meet `sesh`: a terminal app that finds all your LLM coding sessions and lets you read, search, and resume them.
Link below.
This is an unwise statement that can only make people confused about what LLMs can or cannot do. Let me tell you something: Math is NOT about solving this kind of ad hoc optimization problems. Yeah, by scraping available data and then clustering it, LLMs can sometimes solve some very minor math problems. It's an achievement, and I applaud you for that. But let's be honest: this is NOT the REAL Math. Not by 10,000 miles.
REAL Math is about concepts and ideas - things like "schemes" introduced by the great Alexander Grothendieck, who revolutionized algebraic geometry; the Atiyah-Singer Index Theorem; or the Langlands Program, tying together Number Theory, Analysis, Geometry, and Quantum Physics. That's the REAL Math. Can LLMs do that? Of course not.
So, please, STOP confusing people - especially, given the atrocious state of our math education.
LLMs give us great tools, which I appreciate very much. Useful stuff! Go ahead and use them AS TOOLS (just as we use calculators to crunch numbers or cameras to render portraits and landscapes), an enhancement of human abilities, and STOP pretending that LLMs are somehow capable of replicating everything that human beings can do.
In this one area, mathematics, LLMs are no match to human mathematicians. Period. Not to mention many other areas.
Calling on my friend @ericweinstein and @GaryMarcus, who has been one of the few sane expert voices on these matters lately. ๐
h/t @hellheff
@colin_fraser My impression is he either doesnโt know what statisticians actually do, or has a very different definition of โstatisticianโ from me.
@colin_fraser I saw Nate speak during his book tour for The Signal and the Noise. During the Q & A, I asked him about his claim in that book that โmost statisticians are Bayesian,โ citing a tally I did of major stat journals. He replied, โOh, I donโt follow the academic literature.โ
One thing LLMs are really good for is quickly prototyping interactive Shiny apps around well-defined concepts.
I present a Shiny app for exploring Lebesgue integration.
Link to the app below.
R Twitter: is it a known-issue that R's `read.csv()` doesn't support round-tripping from binary64 -> decimal string -> binary64 with 17 decimal digits?
Or is that a bug?
Left is R and right is Python.