today's digest covers new model releases from google, a stark forecast for data center electricity consumption, and practical tools for local ai development. we also look at geopolitical tensions around ai intellectual property and a robotic gripper for data collection.
- google ships three gemini flash models, pro update still missing - google deepmind released gemini 3.6 flash, 3.5 flash-lite, and 3.5 flash cyber, focusing on efficiency and specialized tasks, but the flagship pro model remains delayed.
- data centers to use 4x more electricity by 2035 - bloombergnef forecasts data centers will consume one-fifth of us electricity by 2035, driven by ai compute growth.
- run mythos coding model locally with llama.cpp and pi - a guide to running the qwythos-9b-claude-mythos-5-1m model locally using llama.cpp and connecting it to pi for agentic coding tasks.
- us threatens sanctions on chinese ai models over ip theft - treasury secretary warns of sanctions against chinese ai companies if open source models are found to have stolen american intellectual property.
- grabette: open handheld gripper for robot data - grabette is an open-source handheld gripper that records manipulation tasks without a robot, turning them into training-ready datasets.
other notable stories include ai music now making up over half of daily uploads on deezer, a $34 million funding round for solar installation robots, and a new method to cut memory use in deep network training. for more details, follow the links above.