@InlandCaGuy They released a professionally edited and published version about two months ago.
English
Cullen
643 posts
@JoePostingg I read that in one sitting and it's a rare book I heard of on Twitter and have never seen referenced anywhere else.
That said, boy does it need some edit work
English
everything codex team shipped this week:
- os wide voice dictation
- auto-review mode
- pdf, docs, tex
- sheets, slides
- browser use
- gpt 5.5
enjoy!
English
@london51789 @scaling01 I'm in the US and haven't gotten the update yet - no 5.5 in either web or Codex :/
English
@scaling01 VPN to US and update Codex. You're welcome
English
i guess there are no gb200 gpus in europe
still don't have access to GPT-5.5 on ChatGPT or Codex
English
@ajambrosino How is the remote connections rollout going? That’s still my #1 wanted feature
English
Stop tweeting for a hot minute and update your Codex App to find full browser use, global dictation, non-dev mode, a new auto-review mode that is much safer than yolo, in-app docs and PDF viewer, and ... GPT-5.5.
OpenAI Developers@OpenAIDevs
With GPT-5.5, Codex now gets more of the job done across the browser, files, docs, and your computer. We've expanded browser use so Codex can interact with web apps, and test flows, click through pages, capture screenshots, and iterate on what it sees until it completes the task.
English
@dylanmatt @patwoozey Any concern with giving it docs that have your SSN and other PII in them?
English
Just finished reading Children of Time by Adrian Tchaikovsky
Was a bit more about spiders than I was anticipating, but has to be one of the best contemporary science fiction novels I’ve read😅
Anyone have any other recommendations for recently written sci-fi (say written in the last 20 years)?
English
@woke8yearold I went to DW last month and it was not that expensive. $300/person flights. $500/nt hotel. $150/person/day park tickets. I think a tolerable Tokyo flight itinerary would busy the budget almost on its own.
English
@ClementDelangue llmfit does just that. Nice filterable and sortable TUI based on your hardware.
English
@hanouticelina very cool! is there a simple way just to check what are the latest models that would work on my hardware (I don't ned the working local coding agent)?
English
If you like Claude Code or Codex, you should seriously consider running Agents locally as well! The latest small models (like Qwen 3.5) made this a real before/after moment - and the gap keeps closing. Local coding agents are faster, with more reliable tool calling capabilities, still private, and cost $0 in API bills.
We made it super easy for you to run a local agent with the 𝚊𝚐𝚎𝚗𝚝𝚜 Hugging Face CLI extension - a one-liner that uses 𝚕𝚕𝚖𝚏𝚒𝚝 to detect your hardware and pick the best model and quant, spins up a 𝚕𝚕𝚊𝚖𝚊.𝚌𝚙𝚙 server, and launches Pi (the agent behind OpenClaw 🦞).
One command to find what runs on your hardware and go straight to a working local coding agent!
You should give it a try! 👇
English
You don't need a cloud API for great OCR anymore.
GLM-OCR runs locally with just ~2GB VRAM, handles tables, math equations, and hits ~260 tok/s on a Mac Studio M2 Ultra.
Local models are getting better AND smaller at a crazy pace. If you have a GPU or a Mac, you're already ready for the AI era.
@Zai_org
English
@NielsRogge I appreciate you taking the time. I've watched most of it by now, but I think I'm missing something. I have ggml-org/GLM-OCR-GGUF loaded in LMStudio. No matter what I give it, be it a small PDF or an image of a web page, it yields very little before "EOS Token Found"
English
This is a really cool post about GLM-OCR, but do note there might be better open OCR models on @huggingface!
Check out OlmOCRBench on the hub huggingface.co/datasets/allen…
David Hendrickson@TeksEdge
🚨 Want to parse complex PDFs with SOTA accuracy, 100% locally? 📄🔍 At just 0.9B parameters, you can drop GLM-OCR straight into LM Studio and run it on almost any machine! 🥔 🧠 0.9B total parameters 💾 Runs on < 1.5GB VRAM (or ~1GB quantized!) 💸 Zero API costs 🔒 Total data privacy Desktop document AI is officially here. 💻⚡
English
@Dominik2718 @pashmerepat Do you give it all your docs raw, or redact PII?
English
Currently taxes. And as one would expect, GPT 5.4 has superhuman knowledge of tax regulations and has already found a treaty deduction that I would have never found myself. Codex is just totally unable to fill out the pdfs correctly, as in stuff like putting my name in the right line.
English
@Birdyword Have a Project called Cooking or Home and give it some instructions. Create new threads in that project when relevant. The threads there will share a siloed memory and just those instructions.
English
@lilforestfae Wikipedia says it was created by someone else, and this guy is the second person to be its director/operator.
English
i’m confused, does this guy own the glambot? why is he the executor of the glambot’s will?
Pop Crave@PopCrave
Glambot director Cole Walliser is under fire for his responses to a possible customer inquiring about the camera for a wedding. Walliser reportedly doubted the woman’s ability to pay for his services.
English
@Dimillian Does it automatically make a new git worktree for every thread? I messed with it for a few minutes and had a hard time figuring out what the workflow is supposed to be.
English
@pshemyslaw it's all project for now, I'll probably remove the button on the home it's confusing.
English
@JHWeissmann I spent 8 hours on a bus last month chaperoning a 4th grade field trip. I did enjoy how they sang the Star Spangled Banner leading up to their 6:07 chanting session
English
@parconley Do you have it in the Apple directory? My app of choice (Overcast) can't find it
English
Links!
YouTube: youtube.com/watch?v=8ffN2o…
Spotify: open.spotify.com/episode/6g0nQr…
Should be up on most other platforms as well. Please lmk if not!
YouTube
English
I interview @tylercowen about learning, history, and investing in ideas. Enjoy!
00:00 - Introduction
00:07 - If there were 10 Tyler Cowens
01:10 - Learning curve: ages 14 to late 20s
06:17 - Cracking cultural codes
13:58 - Public intellectuals as investors in ideas
17:16 - Evergreen fields: AI, China, India, Mexico
27:46 - Nuclear war: 70% probability prediction
33:04 - What college students miss
35:58 - Mistakes Gen Z is making
42:12 - Hyperlexia: reading fast as genetic advantage
48:19 - Mentorship
54:27 - Common mistakes mentees make
English
It's sad how much random hate Lex gets.
He's exposed millions of people (who don't have your Twitter feed) to ideas in science/tech which they'd never have encountered otherwise.
And there's 100s of hours of recorded insights from great minds that wouldn't exist without him.
English
@bendreyfuss The Gemini settings should look more like this.
“Thinking” means the new Gemini 3. The image setting on the left means “nano banana” mode is active. Without those both on you are using 6-12 month old AI.
English
How do people get the AI to make whole movies and stuff? I tried to get it to make a very simple thing and it made me go crazy. (Part 1)
English