r3.hyperlogue.tech the r3 product site just launched together with the latest r3 v2.2.0.
r3 is like Claude Artifacts, but focusing on viewing & feedback. It works with any agent harness.
I use r3 in almost all my coding sessions. In the last week, agents created 32 artifacts, published 102 versions, handled 133 feedback threads with 300+ comments.
The use cases include design review, ui prototype, pre-launch changelong revise, launch video review, code base explanation. It's so versatile and deeply integrated into my agentic workflow.
I'd love to hear what you love about claude artifacts or chatgpt site. We can improve r3 to make the features you love available to everyone on every harness!
Rick Ye retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
Nix/NixOS is easily the most underrated tech for AI-native development.
We recently migrated our Tailscale setup to a new account, including 13 VMs, 10 web services, and 7 DNS records. The migration was handled in a single Codex session. My only manual step was running the final deployment script.
Terranix (Terraform in Nix) + NixOS consolidate ACLs, DNS, host configs, and deployments into one repo. That gives AI agents a single source of truth they can actually read, edit, and validate.
The catch? Security patch delivery.
On nixos-unstable this past month:
- PHP critical RCEs: Fixes merged in 75 mins, but took ~78h to reach the channel.
- Tailscale 1.104.4: Landed ~67h after upstream release.
The bottleneck is the sheer scale of the nixpkgs monorepo and (relatively) underpowered CI infra. While Debian or Ubuntu can push isolated security patches to mirrors within hours, a Nix channel bump requires rebuilding and testing thousands of packages before anything moves.
If you run Nix in production, don't rely on upstream channels for 0-days: Spin up your own CI, set up a private binary cache, and use overlays to build hotfix or pin upstream binaries immediately. It’s more involved than an `apt upgrade`, but that’s plumbing you can delegate to an agent anyway.
😂
Ten million people have watched a comedian accidentally destroy the entire TED talk industry.
He filmed the talk once at TEDxNewYork in 2014 and openly admitted he had nothing to say.
Speaking coaches charge $5,000 to teach the exact tricks he used in 6 minutes to say nothing at all.
His name is Will Stephen. He is a comedian from Late Night with Seth Meyers who was invited to give a real TED talk and used it to prove the format is a costume anyone can wear.
Every TED speaker opens with a question, tells a personal story, cites a scientist, and shows a chart. Stephen did all four in six minutes with zero content.
His entire formula fits on a napkin.
Gesture with both hands. Ask the audience a question anyone can answer. Tell a mildly embarrassing story so they say aww. Point to a stranger on a slide and call them important. Flash a chart of random numbers. Wear glasses with no lenses. The audience will think you are brilliant.
That last trick alone has probably cost the public speaking coaching industry a hundred million dollars in sessions clients should have skipped.
"If you watched me on YouTube with the sound off, you would think I was a genius."
That is Will Stephen on the TED stage, three minutes into the talk he was paid to give. He wasn't lying.
Executives pay $15,000 for a keynote coach and still open with a joke the room never laughs at. Founders rehearse their pitch for weeks and still sound like Stephen at minute four, saying "wagga wah, gabba gabba" with authority.
The talk is free on YouTube. The method is written across every TED speaker's playbook for ten years.
Stephen still writes for Seth Meyers. Almost none of the ten million viewers have stopped nodding at empty slides since.
The formula is free. The willingness to actually say something before your next keynote, pitch, or townhall is the entire edge.
Rick Ye retweeted
GPT-6.2 Astra just showed up on the Codex usage page.
The UI still says GPT-6 Astra, but the breakdown flipped to 6.2. A bunch of people posted the same screenshot this afternoon, so it looks like a wide gray rollout, not one account glitching out.
The official model catalog still doesn’t list gpt-6.2-astra. The earlier 6.1 Astra got cut over safety issues, and folks are guessing they fixed it and bumped the number to 6.2.
Tomorrow’s the last day of OpenAI’s 28-day campaign. Wouldn’t be shocked if they announce it then.








