Christopher Webb

Obsessed with what’s next. Profoundly lucky. Bad at running and remembering names. AI Research Engineer at Apple. Opinions are my own.

[Maria Soriano @soriano__maria]

Yes, we have a Foundation Models CLI #WWDC26
- Update your Mac
- Download Xcode 27 (beta)
- run `fm` in Terminal

That's it ✨

[Christopher Webb @conspirator]

Your model should be next!

If you're interested in learning about the latest updates to the Foundation Models framework, you should check out my WWDC session:

WWDC26: Bring an LLM provider to the Foundation Models framework
https://t.co/Qa4DKctYc6

[Christopher Webb @conspirator]

Anthropic and Google recently announced support for Apple's Foundation Models framework, making state-of-the-art Claude and Gemini models available to all Swift developers.

https://t.co/I996DIwhfA

https://t.co/WrolAvqSMG

[Shashank Prasanna @shshnkp]

The force is 💪 with MLX @ #WWDC26
We are super excited to share WWDC26 MLX sessions with you!

For ML engineers:
1/ Using MLX for local agentic workflows with demos showing MLX + open-code @angeloskath
2/ Distributed inference and training on multiple Macs studios with demo showing 1 trillion parameter kimi 2.6

For Swift developers:
3/ NumPy style numerical computing in Swift with MLX

For App developers:
4/ Foundation Models framework API support with MLXLanguageModel package

Link to videos in 🧵

[Matt Shumer @mattshumer_]

https://t.co/ivXRKXJvQg

[Christopher Webb @conspirator]

@usemonologue ios

[Christopher Webb @conspirator]

@jordibruin For some reason my @macwhisperapp app window is ballooning to a massive width which makes it almost impossible to use. Is this just a me thing or are others reporting this as well?

[Awni Hannun @awnihannun]

Check-out the Mac Studio cluster looking good at the Apple NeurIPS booth.

It's running a 1T parameter model using MLX distributed and @exolabs

[Christopher Webb @conspirator]

I’m at NeurIPS this week. Come by the Apple booth to see frontier models running locally with MLX, powering research, coding, and model development.

[Christopher Webb @conspirator]

@skybysoftware @conradev Congrats @AriX and @conradev! I’m really excited for you guys. I look forward to seeing what you cook up at OpenAI!

[Hadi Pouransari @HPouransari]

Introducing Pretraining with Hierarchical Memories:
Separating Knowledge & Reasoning for On-Device LLM Deployment

💡We propose dividing LLM parameters into 1) anchor (always used, capturing commonsense) and 2) memory bank (selected per query, capturing world knowledge). [1/X]🧵

[Christopher Webb @conspirator]

My use case: I generate audio files locally and would love to be able to reference the text while listening.

Again, huge fan! Keep up the amazing work!

[Christopher Webb @conspirator]

@CastroPodcasts I’m a big fan of Castro, especially the Sideload feature. I’m not sure if this has been requested, but I’d love to see you support “Show Notes” and file-specific imagery for Sideloaded files.

[Christopher Webb @conspirator]

@MikeLewis Hey Mike! I was watching the latest Anthropic video, saw the 8-bit character on the desk, wondered what it was, reverse image searched it, and well, here I am. Is this the same little Claw’d from the video?

[Christopher Webb @conspirator]

@pvncher @RayFernando1337 @RepoPrompt ✅

Thanks Ray! Thanks Eric!

[Awni Hannun @awnihannun]

We’re hiring someone to work with us on MLX at Apple. Join the most cracked team to advance the frontier of ML and systems. DMs open.

[Christopher Webb @conspirator]

@BambulabGlobal Linear rail instead of rods? Nice!

[Christopher Webb @conspirator]

Shoutout to those who discovered you can use @Polymaker_3D cardboard spools with @BambulabGlobal refills by just removing the cardboard walls and adding the reusable plastic sides. I swear by Polymaker filament & love my AMS, but respooling/printing adapters is a waste of time!

[Pau Rodríguez @prlz77]

Thrilled to share the latest work from our team at @Apple where we achieve interpretable and fine-grained control of LLMs and Diffusion models via Activation Transport 🔥

📄 https://t.co/TYlwxarrWx
🛠️ https://t.co/gciUcwRNqd

1/9 🧵