Explorer

Being on a podcast I'm a huge fan of is a bit surreal. Grateful to @SamCharrington and @TWIML for the space to talk about why this moment in AI + mathematics feels special -- not just because models are smarter, but because three worlds are finally converging: LLM reasoning, Lean, and code generation. At @axiommathai we are assembling a team of world experts in these three pillars.

@axiommathai : careers@axiommath.ai

Carina Hong

Fri Nov 07 17:16:37

$Here is the plaintext: ### ast-grep vs ripgrep (quick guidance) **Use `ast-grep` when structure matters.** It parses code and matches AST nodes, so results ignore comments/strings, understand syntax, and can **safely rewrite** code. * Refactors/codemods: rename APIs, change import forms, rewrite call sites or variable kinds. * Policy checks: enforce patterns across a repo (`scan` with rules + `test`). * Editor/automation: LSP mode; `--json` output for tooling. **Use `ripgrep` when text is enough.** It’s the fastest way to grep literals/regex across files. * Recon: find strings, TODOs, log lines, config values, or non‑code assets. * Pre-filter: narrow candidate files before a precise pass. **Rule of thumb** * Need correctness over speed, or you’ll **apply changes** → start with `ast-grep`. * Need raw speed or you’re just **hunting text** → start with `rg`. * Often combine: `rg` to shortlist files, then `ast-grep` to match/modify with precision. **Snippets** Find structured code (ignores comments/strings): ```bash ast-grep run -l TypeScript -p 'import $X from "$P"' ``` Codemod (only real `var` declarations become `let`): ```bash ast-grep run -l JavaScript -p 'var $A = $B' -r 'let $A = $B' -U ``` Quick textual hunt: ```bash rg -n 'console\.log\(' -t js ``` Combine speed + precision: ```bash rg -l -t ts 'useQuery\(' | xargs ast-grep run -l TypeScript -p 'useQuery($A)' -r 'useSuspenseQuery($A)' -U ``` **Mental model** * Unit of match: `ast-grep` = node; `rg` = line. * False positives: `ast-grep` low; `rg` depends on your regex. * Rewrites: `ast-grep` first-class; `rg` requires ad‑hoc sed/awk and risks collateral edits.$

Here is the plaintext: ### ast-grep vs ripgrep (quick guidance) Use `ast-grep` when structure matters. It parses code and matches AST nodes, so results ignore comments/strings, understand syntax, and can safely rewrite code. * Refactors/codemods: rename APIs, change import forms, rewrite call sites or variable kinds. * Policy checks: enforce patterns across a repo (`scan` with rules + `test`). * Editor/automation: LSP mode; `--json` output for tooling. Use `ripgrep` when text is enough. It’s the fastest way to grep literals/regex across files. * Recon: find strings, TODOs, log lines, config values, or non‑code assets. * Pre-filter: narrow candidate files before a precise pass. Rule of thumb * Need correctness over speed, or you’ll apply changes → start with `ast-grep`. * Need raw speed or you’re just hunting text → start with `rg`. * Often combine: `rg` to shortlist files, then `ast-grep` to match/modify with precision. Snippets Find structured code (ignores comments/strings): ```bash ast-grep run -l TypeScript -p 'import $X from "$P"' ``` Codemod (only real `var` declarations become `let`): ```bash ast-grep run -l JavaScript -p 'var $A = $B' -r 'let $A = $B' -U ``` Quick textual hunt: ```bash rg -n 'console\.log\(' -t js ``` Combine speed + precision: ```bash rg -l -t ts 'useQuery\(' | xargs ast-grep run -l TypeScript -p 'useQuery($A)' -r 'useSuspenseQuery($A)' -U ``` Mental model * Unit of match: `ast-grep` = node; `rg` = line. * False positives: `ast-grep` low; `rg` depends on your regex. * Rewrites: `ast-grep` first-class; `rg` requires ad‑hoc sed/awk and risks collateral edits.

Former Quant Investor, now building @lumera (formerly called Pastel Network) | My Open Source Projects: https://t.co/9qbOCDlaqM

Jeffrey Emanuel

Fri Nov 07 17:16:32

A useful addendum to your AGENTS dot md or CLAUDE dot md file. First ask codex or claude code to install ast-grep for you if you don't have it already. It's pretty handy for systematically finding general patterns in code that could be tricky to do using regular string matching that's not syntax-aware.

Here is the plaintext: ### ast-grep vs ripgrep (quick guidance) **Use `ast-grep` when structure matters.** It parses code and matches AST nodes, so results ignore comments/strings, understand syntax, and can **safely rewrite** code. * Refactors/codemods: rename APIs, change import forms, rewrite call sites or variable kinds. * Policy checks: enforce patterns across a repo (`scan` with rules + `test`). * Editor/automation: LSP mode; `--json` output for tooling. **Use `ripgrep` when text is enough.** It’s the fastest way to grep literals/regex across files. * Recon: find strings, TODOs, log lines, config values, or non‑code assets. * Pre-filter: narrow candidate files before a precise pass. **Rule of thumb** * Need correctness over speed, or you’ll **apply changes** → start with `ast-grep`. * Need raw speed or you’re just **hunting text** → start with `rg`. * Often combine: `rg` to shortlist files, then `ast-grep` to match/modify with precision. **Snippets** Find structured code (ignores comments/strings): ```bash ast-grep run -l TypeScript -p 'import $X from "$P"' ``` Codemod (only real `var` declarations become `let`): ```bash ast-grep run -l JavaScript -p 'var $A = $B' -r 'let $A = $B' -U ``` Quick textual hunt: ```bash rg -n 'console\.log\(' -t js ``` Combine speed + precision: ```bash rg -l -t ts 'useQuery\(' | xargs ast-grep run -l TypeScript -p 'useQuery($A)' -r 'useSuspenseQuery($A)' -U ``` **Mental model** * Unit of match: `ast-grep` = node; `rg` = line. * False positives: `ast-grep` low; `rg` depends on your regex. * Rewrites: `ast-grep` first-class; `rg` requires ad‑hoc sed/awk and risks collateral edits.

Jeffrey Emanuel

Fri Nov 07 17:16:05

RT @adamdangelo: Enjoyed my conversation with @amasad and @eriktorenberg!

ceo @replit. civilizationist

Amjad Masad

Fri Nov 07 17:14:21

The entire internet being written by ChatGPT makes it feel like you're following a single person with millions of accounts It's the same writer with millions of different human skins Tbh if you went to ChatGPT and just said, "make me a social media feed that I can scroll on that looks like different people", it would pretty much look the same as what you see on a lot of platforms, especially LinkedIn and Reddit (but getting worse on X) I think in general it diminishes the idea sharing value of social media, where you can scroll and hear generally interesting takes that are not "ChatGPT please write me an engaging social media post about [thing]" I strongly believe that unfiltered, full-auto AI needs to be left *off of social media. It's a useful assist, but it shouldn't replace your voice entirely. Our social lives and connections are the last human thing we have left before becoming completely virtually isolated data annotation batteries One can of course argue that social media was already creating virtual isolation but it's not going to get better by replacing human creators with automation Then, there's also something to be said about the potential for social engineering on a massive scale when millions of people are just copy/pasting what ChatGPT tells them and reinforcing the AI's word in public as their own opinions...

Second-time founder 1️⃣ https://t.co/sSfo6UUIdl 2️⃣ https://t.co/LQtr2VGskB Manifold learning, neurosymbolic AI, RL paradigms

Damian Tran

Fri Nov 07 17:13:49

AI can now learn the structures of mathematics and physics, from Collatz sequences to toric Calabi-Yau manifolds. Amazed by this work of @f_charton's at @axiommathai