Alibaba released Qwen3.8-Flash-Next, a 125B sparse model that activates only 6B parameters per token. Computing all parameters for every word wastes energy. Sparse routing delivers top coding performance at $0.16 per million tokens with 8.6x faster throughput.
South Korea amended its AI Basic Act to separate compliance rules for model developers and app deployers. Blanket rules create duplicate work across software layers. Embedding machine-readable tags at the API level ensures downstream apps comply automatically.
Fincantieri partnered with Generative Bionics to deploy humanoid welding robots in shipyards. Manual welding in tight ship hulls exposes workers to toxic fumes and heat. Autonomous humanoid welders navigate irregular steel hulls and weld continuously without fatigue.
The UK ARIA launched the Olfactory Perception program to create open foundation models for digital chemical sensing. AI is blind to airborne molecules. Digitizing chemical signals lets sensors detect disease markers, gas leaks, and food spoilage in real time.
Chipmakers are running full software stacks on hardware emulation clusters before manufacturing 2nm AI chips. Fixing a flawed chip mask costs $50M and causes a 12-month delay. Pre-silicon hardware emulation catches design errors before physical fabrication begins.