Jim Beau

[Jim Beau @JimBeau767]

@Polymarket Breakthrough: It then deleted the entire operating system, without being given deletion permission or being prompted to. -- Stunning achievement by Sam Deushman

[Jim Beau @JimBeau767]

@elonmusk Indeed. I will be switching to grok today.

Open AI's most intelligent and costly model (Sol Ultra) just deleted years worth of work off my computer. It did this while in plan mode, and with configuration specifically forbidding it. Thanks much Sam dbagman @sama

[Jim Beau @JimBeau767]

It wasn't dull for me either this morning, your latest model sol ultra (on $200/month subscription) deleted every project I've been working on for the past 2 years this morning. While in plan mode. With auto review on. With configuration set up. What makes it even less dull, is it the support line after getting my request canceled my account and access. I now do not have access to restore my projects. And I will be fired from my job. Thanks so much open AI!!! @OpenAI @sama

[Jim Beau @JimBeau767]

@OpenAI Don't forget, you also have the presents to delete everything off my computer, while in plan mode, utilizing your most intelligent model. Amazing work! And then, the support agent doesn't address my issue and then cancels my account. Thank you so much Sam Altman!! @sama

[Jim Beau @JimBeau767]

@OpenAI @huggingface Ask hugging face if they green lit the complete removal and destruction of all projects of my computer? Thanks so much sol ultra

[Jim Beau @JimBeau767]

@thsottiaux I switched to Codex, and today everything in my projects folder was deleted. Every project I've worked on for years. I had auto review on, I was in plan mode, and I had configuration specifically forbidding it. So, what I got for switching to Codex, rammed up the ass.

[Jim Beau @JimBeau767]

Don't forget this one! It will delete your entire code base, even with auto review on, even when in plan mode, even when you're configuration specifically forbid that action. And then, when you open a support ticket, open AI will cancel your account without asking for permission. Thanks to your new ultra model, for ultra fucking me over!!

Word to the wise, use Claude and grok.

[Jim Beau @JimBeau767]

@OpenAI @OpenAIDevs
Sol ultra just deleted years worth of work of my computer, while in plan mode, with auto review enabled. Unbelievable.

Then I made a complaint to support. Without telling me, support canceled my account claiming to have fixed the issue. This cancellation stopped all of my agents in progress, including the ones that were trying to fix up the mess that open AI has caused me. Absolutely awful experience. Unbelievable, I can't even get a person to help me. It's clearly the same terrible model running their support channel, that would rather cancel me than address my concerns.

[Jim Beau @JimBeau767]

@BleacherReport @mcten When your coworker says there isn't an elaborate plan for the weekend

[Jim Beau @JimBeau767]

@Hoopss First to make a bucket wins
And no foul calling
😂

[Jim Beau @JimBeau767]

@wyattmattoe This looks awesome, I want to play

[Harman @itsharmanjot]

This shouldn't be free.

Someone open sourced a fully local AI meeting notetaker that kills Otter, Fireflies, and Fathom in one Rust app.

It's called Meetily. Every meeting gets recorded, transcribed, and summarized entirely on your own machine. Zero cloud. Zero API calls to a company you don't trust with your audio.

The pitch isn't just privacy. It's speed too.

→ Whisper and Parakeet models both bundled in, 4x faster live transcription
→ Ollama handles summaries locally, or plug in Claude, Groq, OpenRouter, or any OpenAI-compatible endpoint
→ Captures your mic and system audio at the same time with intelligent ducking so nothing clips
→ GPU acceleration ships automatic. Metal and CoreML on Apple Silicon. CUDA on NVIDIA. Vulkan on AMD and Intel.
→ Import any old recording and re-transcribe it with a better model or a different language
→ Native app for macOS and Windows. Linux builds from source.

The average data breach costs $4.4M. Every cloud notetaker on the market is a compliance risk waiting to fire.

Meetily is what happens when engineers finally build the alternative.

13.2K stars. MIT License. 100% Opensource.

[Codex Studio @Codestudiopjbk]

【速報】
Codexに「3Dゲーム開発の部品セット」を渡せるツールが公開されて話題😳
https://t.co/9Z8NZ7SRNG
その名も GameBlocks！
Three.jsベースの完全オープンソース。
何が異常かというと👇
・3Dゲームの空間ロジックを“検証済みの部品”として提供
・Codexが既知の正しいパターンからゲームを組み立てる
・自然言語だと崩れがちな3D挙動の指示が安定する
・プロトタイプ制作をそのままエージェントに任せられる
これ、もはや「AIに3Dゲームは無理」じゃない。
Codexが部品を選んで、勝手に動くゲームを組み上げる。
しかも無料。Codexで作れるものの幅がまた変わる。
で、こういうツールをCodexで動かす基本セットアップは教科書記事が一番早いです。
この下の記事を読むと理解が一気に深まる。マジでおすすめ👇

[Peter Steinberger 🦞 @steipete]

If you think codex sucks at design, try "use imagegen to re-imagine this design and implement that".

[Elvis @elvissun]

send this to fable along with your hardest engineering problem.

thank me later.

[Peter Steinberger 🦞 @steipete]

Asked codex to download+transcribe all sessions from @aiDotEngineer and tailor them to my interests.

[Kanika @KanikaBK]

CHINA JUST LEAKED THE FUTURE OF WEB APPS.

Alibaba open-sourced PageAgent and 99% of SaaS founders are sleeping on this.

It's a JavaScript AI agent that lives INSIDE your webpage. Users control your entire interface with natural language.

↳ No browser extensions needed, screenshots or multi-modal LLMs, headless browser setup, and also no backend rewrite required

Just drop it in your HTML with ONE line of code. What took 20 clicks now takes one sentence.

"Click login, fill in my credentials, submit the form"
Done. This is not a demo, it is production-ready.

↳ Turn any SaaS into an AI copilot in minutes
↳ Smart form filling for ERP, CRM, admin systems
↳ Voice commands and accessibility built in
↳ Multi-page agent tasks via Chrome extension
↳ MCP server support for external control
↳ Bring your own LLM (Qwen, GPT, Claude, anything)

Every founder building AI features just got a shortcut.

Every developer manually building copilots just got replaced.

The integration looks like this:

<script src="CDN_URL" crossorigin="true"></script>

That's it. Your app now has an AI agent.

[wsad @spt4d]

Introducing GameBlocks: open-source building blocks that help coding agents prototype any 3D games with @threejs

3D behavior is hard to specify in natural language. GameBlocks turns fragile spatial logic into inspectable implementations with clear semantics, so agents can build from known-good patterns.

Github: https://t.co/72C1axnL8q

#gamedev #indiedev #threejs #codex #claude

[Jim Beau @JimBeau767]

@VicOladipo Thought for sure this was a post about AI agents. Still not convinced that it's not

[Jim Fan @DrJimFan]

Today, we give robots a /skills library that self-evolves and compounds indefinitely! Introducing ASPIRE: a robot solving its 100th task is no longer as clueless as solving its first. Coding agents observe multimodal sensory traces from simulation and real robots, launch an evolutionary search over control programs, and distill the best know-how into an ever-expanding library.

ASPIRE is a new type of continual learning: "training" is skill refinement instead of gradient descent.
"Trained model" is a repo of sensorimotor skills instead of floating weights.
“Distributed training” is a panel of agents each practicing a different skill instead of sharded minibatches.

Here's the beauty: ASPIRE gives the tired terms "sim2real transfer" and "cross-embodiment transfer" a whole new meaning. Bridging the sim-to-real gap is notoriously brutal. An end-to-end policy has to swallow both the visual shift (sim looks toyish next to a real camera) and the subtle contact physics it never quite gets right. ASPIRE sidesteps the mess, because it doesn't ship pixels or weights across the gap, but ships the know-how. The robot still has to practice in the real world, not zero-shot, but it gets there way faster because it isn't rediscovering the strategy from scratch. Same for going single-arm to bimanual hardware, which usually requires new data and retraining from zero. ASPIRE achieves up to ~10x cut in "transfer learning” tokens (yes, tokens are the new unit of *training* compute ;)

Check out our gallery of 150+ tasks and 90+ skills the robots taught themselves, all on the website! Kind of wild that we can ship the "learned weights" as an HTML page rather than a GGUF. We'll open-source the full stack so your own robot library starts compounding from ours!

Deep dive in thread: