OPEN SOURCE - AI Dynamics

New Frontier Open Source Lab to Launch This Year

By

–

27 March 2026 18h56

A frontier opensource lab in the West will be born this year. Zero doubt. It requires serious capital, like I’ve said before. Working on it. One day I’ll tell the story of how it started in a basement and ended at the frontier. x.com/vSouthvPawv/st…

→ View original post on X — @theahmadosman

27 March 2026

SAM 3.1: Object Multiplexing Enhances Video Processing Efficiency

By

@aiatmeta

–

27 March 2026 18h26

We’re releasing SAM 3.1: a drop-in update to SAM 3 that introduces object multiplexing to significantly improve video processing efficiency without sacrificing accuracy. We’re sharing this update with the community to help make high-performance applications feasible on smaller,

→ View original post on X — @aiatmeta

27 March 2026

Claude Code with HF Papers CLI for Autoresearch

By

@_akhaliq

–

27 March 2026 17h54

Claude code using HF papers CLI to do autoresearch

→ View original post on X — @_akhaliq

27 March 2026

Qwen 3.5 9B Model Now Available on Ollama

By

@datachaz

–

27 March 2026 17h37

this one → https://
ollama.com/library/qwen3.
5:9b
…

→ View original post on X — @datachaz

27 March 2026

Z AI Releases GLM-5.1, Open-Source Expected Soon

By

@testingcatalog

–

27 March 2026 17h14

Z AI released GLM-5.1 and made it available to all GLM Coding Plan users. It is also expected to be open-sourced in the first half of April.

→ View original post on X — @testingcatalog

27 March 2026

TensorFlow natively runs on MacBook Pro M4 finally

By

@tunguz

–

27 March 2026 16h16

I just installed TensorFlow nightly 2.22.0-dev20260326 on my MacPro Max M4 with Python 3.13. Almost a year after I bought it I can *finally* run TensorFlow natively on the bare metal on this machine.

→ View original post on X — @tunguz

27 March 2026

Project N.O.M.A.D: Running Local AI for Offline Knowledge Management

By

@alphasignalai

–

27 March 2026 16h01

Someone open-sourced a survival computer that works even when the internet goes down. Project N.O.M.A.D turns any Linux machine into a fully offline knowledge server. No internet needed after setup. The AI runs through Ollama. You can chat, write, and code locally. Upload

→ View original post on X — @alphasignalai

27 March 2026

Mojo Kernels: Reducing conv2d Code from 870 to 130 Lines

By

@jeremyphoward

–

27 March 2026 16h00

130 lines instead of 870. That's the difference between our conv2d implementation on Blackwell and CUTLASS's. We broke kernels into three swappable pieces: one for moving data, one for coordinating the pipeline, one for compute. When you need a new kernel, you only change the piece that actually needs to change. Part 3 of our Structured Mojo Kernels series walks through the details: modular.com/blog/structured-…

→ View original post on X — @jeremyphoward, 2026-03-27 15:00 UTC

27 March 2026