What's trending in AI and technology on X and YouTube — updated twice daily
Last updated: March 30, 2026 at 8:02 AM EDT
⏳ Loading audio…
Local AI can be so good, but you'd need about 12k USD to get it. Then it's not so great Here's a Q4 of Qwen3.5-262B-REAP Weights 131GB KVCache 50GB 256,000 context 350 tokens/s prefill warmup 4,000 tokens/s prefill cache 36 tokens/s generation Vision enabled REAP is good
This is god tier and fits on 24gb + It crushes everything up to 6x its size and ties with Gemini deep think and Deepseek on Math
I will get stuck into tuning and testing Tinygrad Mac M3 and RTX Pro 6000 Workstation 300w edition and a Razer eGPU TB5 edition. @__tinygrad__ Lets see how we go.
Episode 201: Fracking Apple Silicon We plan to connect millions of Apple silicon chips into the world's largest network for agentic compute.
Naive RAG vs. Agentic RAG, explained visually: Naive RAG has well-known failure modes: - It retrieves once and generates once. If the context isn't relevant, it can't search again. - It treats every query the same. A simple lookup and a complex multi-hop reasoning task go through the same pipeline — no adaptation.
When the DOGE experiment launched last year, it was framed as the ultimate solution to government bloat.
When a digital fruit fly moves, what are we really witnessing?
Watch now | "The best time to start," Tom told me, "was yesterday."
Get AI & tech trending updates delivered to your podcast app
📻 Subscribe via RSS