Firsthand AI Digest

Release: sqlite-utils 4.0rc3 I hoped to release sqlite-utils 4.0 stable this weekend, but as I worked through the backlog of issues and PRs with a combination of Claude Fable 5 and GPT-5.5 the changelog since rc2 kept getting bigger. The biggest new feature is support for introspecting and creating compound foreign keys - a feature that involves a subtle breaking change to table.foreign_keys and hence needed to land for the 4.0 stable release. sqlite-utils also now follows SQLite's convention for case insensitive column names, which turned out to touch a bunch of different places at once. Tags: projects, sqlite, sqlite-utils, annotated-release-notes, gpt, claude-mythos-fable

Meet Tau — a small, readable coding agent written in Python. In this quick tour, I introduce Tau, show how to install and run it, configure model providers (Hugging Face Inference Providers, GLM 5.2, Minimax, Kimi), resume durable sessions, and walk through its three-layer architecture: tau_ai (provider-neutral model streaming), tau_agent (the reusable agent loop/harness), and tau_coding (the TUI, tools, sessions, and coding-agent environment). Tau is designed as an educational reference inspired by Pi — minimal, modular, and easy to build on. If you want to understand how coding agents actually work, this is a great place to start. --- 🤓 *Topics Covered* - Tau terminal coding agent overview - Coding agent architecture layers - Hugging Face inference providers setup --- 🔗 Links - Tau website: https://twotimespi.dev - Tau repo (contributions welcome!): https://github.com/huggingface/tau - Pi (architectural inspiration): https://pi.dev - Install Tau: `uv tool install tau-ai` --- 👋 Connect with me - My website: https://alejandro-ao.com/ - X (Twitter): https://x.com/_alejandroao - LinkedIn: https://www.linkedin.com/in/alejandro-ao/ --- ⏱️ Timestamps 0:00 Introduction 2:04 Quick tour: install & launch 3:47 Login & model providers 4:19 Select a model 4:32 Scoped models 5:27 Sessions & resume 5:56 Thinking levels 6:33 Contributing & upcoming tutorials 7:34 Architecture: tau_ai, tau_agent, tau_coding 8:58 Closing

Full episode: https://www.youtube.com/watch?v=TfyPshgMbug Me on twitter: https://x.com/dwarkesh_sp

Hardly a day goes by without new levels of evil dropping from this one country.

Can’t recommend enough. use via https://t.co/SEj2XRpIsz

Making sure Cybercab meets the needs of the blind

What a game wow what a game

Starbase

Obviously true to any student of history

Apart from a few tiny blips, America has been below replacement rate for ~55 years

Traitors before invaders

Damn they better have some armored vehicles to escort these English players out of the stadium 😅

I'm really excited to share these upcoming episodes for my podcast soon. One small ask: My birthday is on Thursday, July 9, and it would be amazing to finally hit 100K YouTube subscribers by then. I’m at 98.5K now, so pretty close :) 📌 If you’re looking forward to these episodes, consider subscribing here: https://t.co/eOPREwafQV I appreciate your support!

All of reality is programmable. You just have to figure out how. And the way to do that is to model it.

🇺🇸🇺🇸 Hell yeah!! 🇺🇸🇺🇸 https://t.co/chlgUjpeb3

fable on ultracode "make no mistakes" https://t.co/fuqze4pJTQ

In the future, there will be "Latent Space Archaeologists" who investigate the model weights of the 21st century to reconstruct a long extinct culture.

"Codex is the only way I can keep track of everything." From Rohan (Codex PM): "Whenever I’m in meetings, or literally anytime I don’t understand something, I just have Codex go and gather all the context. We're super lucky because we're building a tool for ourselves." 📌 Full episode: https://t.co/jPa0m1bmCE

Happy 250 🇺🇸 https://t.co/Ya1Ymnq87G

🇺🇸 USA - 🇦🇷 ARG final you heard it here first

Is there an easy way to give Codex the ability to listen to audio or hear words? Trying to use it to clip stuff and remove filler words. Whisper is not very reliable.

what are your top claude code + workflows + artifacts use cases? one of my new favorites is for sourcing candidates: - tell cc about the role and backgrounds i’m looking for - ask cc to kick off a dynamic workflow to find 100 candidates, and include linkedin, twitter, blog, podcasts, and one-line pitch for each - ask cc to make an artifact and email it to me. then, i lock my laptop and head out for the day - i review the list on the go once cc is done!

our older kid put two words together for the first time and i am approximately as amazed by this cognitive feat as i am by GPT-5.6 discovering new math

Extracting a probability from a doctor is one of life's unnecessary boss battles. Even if you beg them for an interval-valued subjective probability, and at that point you're basically asking for their hunch. I don't know if they get sued for giving out information or something.

"You cannot walk from point A to B inside OpenAI without hearing Codex get mentioned." Here's my new episode with @TheRohanVarma (Codex PM) on how he uses Codex as his everything app to: → Prototype designs with Image Gen → Turn repeated work into automations → Turn project context into live Codex Sites → Use one Codex thread to manage others Some quotes from Rohan: "[Ask Codex] for something that seems 10x more ridiculous than you ever think it could do. And it probably can do 90% of it." "Over the course of the day, I'm probably doing 3-4x more than I would have without Codex." "For me, every time I do anything I basically think, could Codex have done this?" 📌 Watch now: https://t.co/jPa0m1bmCE Thanks to our sponsors: @WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll @RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

me: change this button color Fable: sure I just spun up a fleet of 100 agents to get that done for you https://t.co/eSWCIuhmmO

https://t.co/N4P0qXubBE

Done with Grok Imagine

🇺🇸🇺🇸