Worth noting

the pressure of FOMO

🔗arxiv.org/abs/2506.02153

It’s coming 🤔

Introducing DeepSeek-V3.1: our first step toward the agent era! 🚀 🧠 Hybrid inference: Think & Non-Think — one model, two modes ⚡️ Faster thinking: DeepSeek-V3.1-Think reaches answers in less time vs. DeepSeek-R1-0528 🛠️ Stronger agent skills: Post-training boosts tool use and

Re-upping this AI startup news from Friday 👀 Meta and Midjourney held partnership talks including about a potential acquisition this summer, joining Meta convos with Runway, Pika, Higgsfield, HeyGen and Krea (the later two unreported). Shortly after we published at

Workflow for this one is nano-banana to Wan 2.2 (on KREA AI). Asked nano-banana to turn the pets into gangsters in an alleyway with grills and chains - it took creative liberty with the hat 😂 Prompt for Wan 2.2 was "light animation of dog and cat gangsters."

In era of pretraining, what mattered was internet text. You'd primarily want a large, diverse, high quality collection of internet documents to learn from. In era of supervised finetuning, it was conversations. Contract workers are hired to create answers for questions, a bit

pretty crazy that people are vibe coding music now (this is dj_dave) she uses strudel a new live coding app to write music in your browser (open source too) i've never knew this was possible before

sf has the most autistic billboards

Wow Long-distance world exploration is amazing

Tried the HunyuanWorld-Voyager and found it very impressive. A scalable “world memory” system maintains geometric stability and coherence across any camera movement. It‘s the world’s first ultra-long-range world model with native 3D reconstruction, redefining AI-driven spatial

It's been about a year since my team has fully adopted all the AI coding tools (Cursor, Claude Code) And day to day I am feeling the added cruft in the code base. Unit tests are not catching regressions. Unneeded mocking, comments, are left in between. More refactoring is needed

wow amazing 🤩