WEEKLY DIGEST
Week in AI: Jun 01 — Jun 05, 2026
The top stories from our coverage this week.
Autonomous defense manufacturer Mach Industries raises $300M for government contracts and product development.
https://www.therobotreport.com/autonomous-defense-manufacturer-mach-industries-raises-300m/
Tesla faces a robotics rival in China as sales recovery gains momentum. @tesla
https://news.google.com/rss/articles/CBMisAFBVV95cUxQa0pKcjlvemtuNkNTbHlWWXUyeUlBS241U1FSbWZ5ZmV2U1JaRmUyMWNXZ0loUEsyLWlvUTE1bmJSSWFDOGEyeUM0V0gxX0Z6bnREQ3ZQQ0xqclpOcTBJWGwtQUFCZi13ZkhhaHJXRjNYWnJHLTJfNlF6ZXUxZWJ1b2hieURJcnZZSlNEZ2RvWDdDXzYzVmVqclVnQV9rZDFzc2FOU3lNel9pV2xTSHNCeA?oc=5
Tesla, Figure AI, Apptronik develop humanoid robots for factory work. @Tesla @FigureAI @Apptronik https://news.google.com/rss/articles/CBMinAFBVV95cUxPaEJ2SF9xTU5kaml5UFljamlEelhiVnJsdV9KU0dvMGJhMWRnUjY3OXZ4Zkd5Z0hxdEdOZzZXaVZaOTgwVjBJNHEyUDZxY19nMzJHSnR6VEtzSk1iZ1VwTG9lc1l1clNJZl9OdmZGelpjN3pUaGQ1ZENpZlRleE9TMHFhQ3ZDajJhUUJwRVpmVE44VjZfRjZJTHdHazI?oc=5
Generalist raises $400M to scale general-purpose AI models, boosting task success to 99%
https://www.therobotreport.com/generalist-raises-400m-to-scale-its-general-purpose-ai-models/
NVIDIA releases Nemotron 3 Ultra: 550B open Mixture-of-Experts hybrid Mamba-Transformer for long-running agents. @nvidia
https://www.marktechpost.com/2026/06/04/nvidia-ai-releases-nemotron-3-ultra-an-open-550b-mixture-of-experts-hybrid-mamba-transformer-for-long-running-agents/
<post>
<line1>Microsoft and @Nvidia partner on AI PCs running agents next week at Computex and Build</line1>
<line2>https://the-decoder.com/microsoft-and-nvidia-reportedly-team-up-on-ai-pcs-that-run-actual-agents-instead-of-copilot/</line2>
</post>
<end>
Answer:
New research audits open skill ecosystems to measure how structured workflows impact large language model agent performance.
https://huggingface.co/papers/2605.23657
@JeffDean details ML systems for trillion trillion FLOPs in new technical overview.
https://www.youtube.com/watch?v=139UPjoq7Kw
LURE introduces Live-Usage Replay Evaluations to mitigate frontier models' evaluation-awareness risks, enhancing safety benchmarks.
https://www.lesswrong.com/posts/WKuGzrtCnAAArjj2N/lure-alignment-evaluations-to-reduce-evaluation-awareness
NVIDIA releases open-source physical AI tools, Isaac GR00T humanoid robot. @NVIDIA https://www.therobotreport.com/nvidia-releases-new-updated-tools-physical-ai-gtc-taipei-computex/