AI & Tech Developments - Apr 01

If you have a Thunderbolt or USB4 eGPU and a Mac, today is the day you've been waiting for! Apple finally approved our driver for both AMD and NVIDIA. It's so easy to install now a Qwen could do it, then it can run that Qwen...

— @__tinygrad__

For those wanting an Opus distill but in a smaller, more accessible 9B size that easily fits on GPUs with 16GB VRAM or less: Jackrong just dropped Qwopus3.5-9B-v3 — a fast-iterated reasoning-enhanced model based on Qwen3.5-9B with Claude Opus-style patterns distilled in.

— @outsource_

Iran declares Butlerian Jihad

— @CarlZha

The biggest reduction in KV cache memory comes not from quantization or MLA, but from latent compaction, along the sequence dimension. More strong results coming soon with Attention Matching.

— @AdamZweiger

Claude Code 2.1.89 has been released. 9 flag changes, 52 CLI changes Highlights: • PreToolUse hooks can return 'defer' to pause headless tool calls; resume with -p --resume to re-eval • PermissionDenied hook fires after auto-mode classifier denials; return {retry:true} to let

— @ClaudeCodeLog

A solo dev rebuilt Google’s new algorithm with Claude in 7 days, made it 3.7x faster, and got a 35B model running on a MacBook with 4.6x compressed cache. Google published the paper. He shipped the code.

— @RoundtableSpace

working on an ai analysis model specifically targeted at shorter text for social media posts browser extension soon

— @liquiditygoblin

Apple approves driver for eGPU on Mac, enhancing performance for AMD and NVIDIA users.

— @__tinygrad

New model Qwopus3.5-9B-v3 released, optimized for smaller GPUs.

— @outsource_

Insights on KV cache memory reduction through latent compaction.

— @AdamZweiger

Announcement of Claude Code 2.1.89 release with significant updates.

— @ClaudeCodeLog

Solo developer rebuilds Google's algorithm with Claude, achieving remarkable speed.

— @RoundtableSpace

Discussion on new AI analysis model for social media posts.

— @liquiditygoblin

Just hard to overstate the scale of this $122B raise for OpenAI.

— @RyanWatkins_

Bubbles tend to peak at ~40% concentration levels. The current AI bubble is at 41%.

— @ekwufinance