Mamba-style architectures offer linear scaling with sequence length, challenging the transformer monopoly on language modeling.
A research team observes unexpected cooperative strategies and communication patterns emerging in autonomous multi-agent systems.
The MCP standard for AI tool use is being adopted by major platforms, promising interoperability across AI systems and services.
New open-weight models from the community are closing the gap with proprietary systems across reasoning, coding, and instruction following.
Ollama adds structured output, native tool calling, and improved vision model support in its biggest release yet.
A Stanford study finds that developers using AI assistants produce code with 40% more security vulnerabilities than those coding without AI help.