Pinned Tweet
What AI will become… random thoughts on music, movies, games…
Music 🎶
Extend your favourite songs… have a fav song that isn’t long enough. Ask AI to extend it more. Love a certain part? Make it longer
Extend an album. 💿
Continuously make new songs in the style and feel
Steve GrAIhAIm retweeted
We’re opening up access to Atlas to everyone today.
We’ve built apps and API endpoints to help you use Atlas any way you can imagine: bring your ideas for building a 3D world and come away with keyframes, video clips, or 3D Gaussian splats.
Steve GrAIhAIm retweeted
i used fable 5.2 to make this "hand-painted" game in less than an hour
Every frame is about 1.5 million brush strokes, running at 200 fps in a browser tab. Open source too!
Steve GrAIhAIm retweeted
for the past few weeks, I’ve been exploring painting with Fable 5.2 in code. one of the most common critiques I heard about Opus 5.5’s paintings was that it was only emulating real art. I thought that was fair, so i’ve been asking Fable 5.2 to develop personas—artists with some sort of a back story and a consistent style roughly identifiable in time
not knowing what to expect, i’ve watched this project evolve in directions i never predicted. I give comments or steers, and Claude updates or revises the engines that power the brushes and paint. where paintings come up i like or find intriguing, i ask it to go deeper. these are some favorites.
each of these is a python program, some one to two thousand lines of code, that places every stroke through a simulated paint engine with different bristles. most take a few thousand strokes, but i’ve pushed much higher with varying results. Claude does not use an image model, nor render or trace in advance
Now in beta: Businesses in ChatGPT.
Claim your business, manage your information, and get insights on how people engage with you.
Eligible businesses can also build a Business Agent to answer customers’ questions and help them explore what you offer.
openai.com/index/introducing…
Steve GrAIhAIm retweeted
You can now embed your website in Claude and have it be interactable. You can do this via MCP apps, where we've enabled frame_domains support for verified plugins.
It was a blast to build this out with the team & our partners, let us know what you think!
Replying to @claudeai
Also new: with interactive connectors, you can now open the full version of a tool inside Claude. Work in Adobe, Atlassian, and Box beside your conversation. Make changes yourself, or ask Claude to make them with you.
Rolling out on web and desktop.
Steve GrAIhAIm retweeted
Entire prompt:
'You know, you should also build me a FPS. Insanely detailed quality pls, something that looks visually like it can keep up with AAA. Make it small in scope (like a tiny level) but otherwise impressive photorealistic graphics.'
Fable 5.2 one shot it.
I wanted to test Fable 5.2's quality with one-shot prompts without human feedback.
Everything in this game is from scratch: models, sounds, textures. ~$550 at API pricing, ~2.7M output and 445M input tokens, heavily cached. Took 8 hours!
Play it here: claude.ai/artifact/BGY1SztrA…
Vibecoders, 𝗜'𝗺 𝗯𝗲𝗴𝗴𝗶𝗻𝗴 𝘆𝗼𝘂, before your next website or app, paste this into Claude Code or Codex:
"I want this to be as perfect as Linear's website or Apple Notes: simple purpose, simple architecture, simple design, simple implementation. Cut taglines, helper text and secondary labels. If a screen needs them to make sense, fix the layout. Every decision should elevate the user experience (UX), developer experience (DX) and agent experience (AX) at once, without breaking anything."
Steve GrAIhAIm retweeted
Got fable to hand-write weights of a 100k param model to recognize real versus fake legal citations.
Basically one-shotted a full NN that’s better than training it
Steve GrAIhAIm retweeted
There's a buzzy new paper about GPT-6 Astra being able to directly interpret model weights (prompts full of floating point numbers) - I tried it with Qwen 3.8 27B on my laptop and found that, after 45+ minutes, that model could achieve the same thing simonwillison.net/2026/Nov/3…
Steve GrAIhAIm retweeted
i remember back in the old days of summer 2025, when i wrote a blog post about how claude code calls haiku for things like those whimsical words while it works, draining my sub.
never change, ant.
gist.github.com/flamerix-ai/…
Steve GrAIhAIm retweeted
Wanted to share a method for information discovery with Codex App I've been using for the past 4 months that has brough me a lot of value, that is not talked about much docs.google.com/document/d/1…
Steve GrAIhAIm retweeted
My friends were all talking about prepping so I made a survival LLM. A free app that runs an AI mdoel plus 1.5 million offline articles (medical guides, FEMA, survival manuals, wikipedia) entirely on your computer. No internet needed.
brangus.io/survival-llm
Steve GrAIhAIm retweeted
one thing that made this click for me — with decision models, you have to define categories and it returns probabilities that the input matches those categories.
but so much of the data we interact with is completely new to us. how can you define categories for something you've never read?
with clairic, you can extract values directly from the inputs, without any premeditated categories. the pipeline becomes clairic to extract values → clef to make decisions with those values → LLMs to take actions on those decisions. all of these models run on Workers AI.
a brand new model paradigm! introducing 𝚌𝚕𝚊𝚒𝚛𝚒𝚌: a system one model for extracting, computing, estimating, and predicting values from unstructured inputs.
blog.cloudflare.com/clairic-…
Steve GrAIhAIm retweeted
This is indeed super easy, I highly recommend trying it
There's no "hack" it's really just adb commands through your local network which Claude knows by heart
I tried it and disabled Xiaomi's analytics and statistics packages to save a nice 131 MB of RAM (out of less than 2 GB)
Xiaomi doesn't really have much bloatware so I'll probably leave it at that
I was going to create a GitHub project to make it easy to do what Ryan did but it's so easy that Haiku can probably do it for under $0.10 (Sonnet did it for under $0.50)
Here is the full prompt:
"Tell me what information you need to access my smart TV through adb. Save a backup using `db shell pm list packages` then sort the packages into likely bloat. I will then tell you which packages we can then disable using `pm disable-user --user 0 <package>`. If anything goes wrong, we can revert this with `pm enable <package>`."
Hacked my TV with Claude Code today
No more ads, bloatware, spyware or laggy UI
I also got a beautiful screensaver with my revenue, calendar, weather, and notifications for new Stripe payments
It was super easy! I sent it the listing for my tv, asked "can you hack this?" And it talked me through how
Steve GrAIhAIm retweeted
A fascinating visualization of Vivaldi’s “Winter” that makes the music feel more intense. Almost like seeing the sound itself.
credit Gilmar Araújo - Conde X
Steve GrAIhAIm retweeted
With everyone busily using LLMs + Blender/Three.js to make 3D models, I thought I'd try a different geometric representation.
This dragon isn't a mesh, NeRF, 3DGS, or video model: it's 27kb of GLSL code (generated via LLM), defining a closed-form implicit surface. 🧵 [1/n]
Steve GrAIhAIm retweeted
VERY underrated classification AI flow:
ALL DONE IN CLAUDE CODE:
1) Have fable create a prompt for classifying a thing
2) Then, have claude code run 100 samples through with fable as the model. Spot check these. These are source of truth.
3) Then, have claude code build the wiring to run it through jev first.
4) Have fable examine the confidence scores of jev. You're looking to see 'if confidence falls off <.7, should we fall back to the frontier?'
5) Have claude code string up a fall back to luna when confidence from jev is < x.
6) Evaluate the classifications jev/luna made using fable.
7) Have fable update the prompt to catch edge cases
8) Repeat until good
You now have a low cost VERY accurate classifier for almost anything
Steve GrAIhAIm retweeted
Claude Managed Agents dynamic workflows are now available in public beta. It's a new type of multiagent orchestration, built for your most ambitious workloads.
A lead agent writes a plan that runs across many agents in phases, combining the results at the end.
So, how does that feel? Well, it feels like we'd been searching the woods for a missing hiker for 4 days, really bonding as a team, getting great at reading maps, and then a helicopter showed up, found him in ten minutes, handed him to us alive, and flew away. So, I guess I'd say, "not great".
41
Steve GrAIhAIm retweeted
Or give EmDash a try, it’s V1 as of this week!
blog.cloudflare.com/emdash-c…


































