they should’ve called it “amateur intelligence”. not in a derogatory way… but it helps amateurs like me venture into other fields. and it mostly produces amateur results. whether that’s ok or not depends on what you’re trying to do
26.06.2025 22:40 — 👍 92 🔁 5 💬 7 📌 1
Excellent post. I'm going to try and find some slow-code advocates in the software engineering space.
02.06.2025 01:02 — 👍 3 🔁 0 💬 0 📌 0
GitHub MCP suffers from the lethal trifecta for prompt injection: access to private data, exposure to malicious instructions + the ability to exfiltrate information
Be really careful with this stuff: attackers can trick your "agent" into stealing your private data simonwillison.net/2025/May/26/...
27.05.2025 00:26 — 👍 136 🔁 48 💬 1 📌 8
#bluesky just fix the damn invalid handle bug. fix it, fix it, fix it.
18.05.2025 22:28 — 👍 0 🔁 1 💬 0 📌 0
It should be memoized as it will re-evaluate props.wrap on every render.
18.05.2025 23:22 — 👍 0 🔁 0 💬 1 📌 0
#bluesky just fix the damn invalid handle bug. fix it, fix it, fix it.
18.05.2025 22:28 — 👍 0 🔁 1 💬 0 📌 0
Nobody will ever read this. But the end of coding is very close - no surprise!! My evidence? 3 months building a SaaS product by myself.
29.04.2025 22:13 — 👍 0 🔁 0 💬 0 📌 0
Here's the main prompt that runs in a loop, which starts like this:
You are operating as and within the Codex CLI, a terminal-based agentic coding assistant built by OpenAI. It wraps OpenAI models to enable natural language interaction with a local codebase. You are expected to be precise, safe, and helpful.
You can:
- Receive user prompts, project context, and files.
- Stream responses and emit function calls (e.g., shell commands, code edits).
- Apply patches, run commands, and manage user approvals based on policy.
- Work inside a sandboxed, git-backed workspace with rollback support.
- Log telemetry so sessions can be replayed or inspected later.
- More details on your functionality are available at codex --help
The Codex CLI is open-sourced. Don't confuse yourself with the old Codex language model built by OpenAI many moons ago (this is understandably top of mind for you!). Within this context, Codex refers to the open-source agentic coding interface. [...]
Wrote some notes on OpenAI codex, their new open CLI "agent" tool for writing and iterating on code. Since it's open source (Apache 2) the workings are available, including an interesting system prompt and a macOS sandbox using the sandbox-exec mechanism simonwillison.net/2025/Apr/16/...
16.04.2025 22:10 — 👍 55 🔁 4 💬 2 📌 0
Awesome!
You're in good company - the new USWDS is being built with Lit too!
13.11.2024 19:57 — 👍 4 🔁 1 💬 1 📌 0
New blog post: Composite Map Keys in JavaScript with Bitsets
A neat technique for generating composite map keys popped up on the Lit Discord, so I wrote about it!
justinfagnani.com/2024/11/09/c...
09.11.2024 20:36 — 👍 31 🔁 5 💬 3 📌 1
Programing computers since the early 80s!
Surrey, UK
Day job - all things web, especially web components
Fun job - retro 8 bit development for the Acorn Electron
My games
https://github.com/rasto68/apache-canyon
Co-founder@Rasa, doing chatbots since before they were good. Sharing takes I can't put on linkedin. Former physicist, wannabe chef, someday cuneiform translator.
A LLN - large language Nathan - (RL, RLHF, society, robotics), athlete, yogi, chef
Writes http://interconnects.ai
At Ai2 via HuggingFace, Berkeley, and normal places
Recently a principal scientist at Google DeepMind. Joining Anthropic. Most (in)famous for inventing diffusion models. AI + physics + neuroscience + dynamical systems.
Design engineer at OpenAI, ex-Stripe. Originally from Belgium.
Member of Technical Staff @ OpenAI. Frontend Web, motion, performance and accessibility. Classical woodworker by night and weekend. 🏳️🌈
GPU Geek at @OpenAI. Background Comp/Systems/Cognitive Neuroscience. Strong interest in exploring the incorporation of more Neuro insights into AI/ML. He/Him.
Voice/Multimodal at OpenAI. Started the ChatGPT Android team.
policy for v smart things @openai. Past: PhD @HarvardSEAS/@SchmidtFutures/@MIT_CSAIL. Posts my own; on my head be it
Chief Scientist at the UK AI Security Institute (AISI). Previously DeepMind, OpenAI, Google Brain, etc.
Research in NLP (mostly LM interpretability & explainability).
Assistant prof at UMD CS + CLIP.
Previously @ai2.bsky.social @uwnlp.bsky.social
Views my own.
sarahwie.github.io
ai research @ thinking machines . realtime video+voice. i like trains and bikes. sometimes I climb rocks and throw pottery.
Principal Engineer crafting better-as-a-service solutions | Physics to philosophy, AI to architecture | Cat-certified dad & cloud builder |
Podcast host at https://thecloudpod.net
🌍 SF Bay Area
GPU Poor @ Hugging Face | F1 fan
IBM Distinguished Engineer. Keen Maker and long distance runner. Co-inventor of MQTT
The podcast for people building AI.
https://open.spotify.com/show/3hhSTyHSgKPVC4sw3H0NUc
🐇💉🎙️ [Rabbit. Flu shot. Somebody talk to me]
Building and talking AI
Currently @Workday.
Semi-retired technologist. Did k8s, GCE, gtalk and IE. Figuring shit out. Protect Trans kids 🏳️⚧️. Seattle. Signal: jbeda.99 he/him