Category: AI
OpenAI’s Navier-Stokes Proof Raises Questions About AI and Mathematical Understanding
OpenAI’s reported Navier-Stokes proof remains under review, prompting mathematicians to question whether AI can solve problems without explaining the ideas behind them.
Voice AI’s Harder Benchmark Is Local Understanding, Not Fluent Speech
Voice agents increasingly sound human, but accents, idioms and cultural expectations expose gaps that multilingual support alone cannot solve.
Nvidia launches open-source platform for controlling autonomous AI agents
Nvidia has launched an open-source platform that applies safety controls across AI agents, runtime software, compute infrastructure and hardware.
NVIDIA Releases OpenShell 0.1.0 to Enforce Runtime Policies for AI Agents
NVIDIA’s OpenShell 0.1.0 adds sandboxing, credential controls and policy enforcement around AI agents without requiring changes to their workloads.
Anthropic CEO to Skip Australian Senate AI Hearing as Oversight Questions Grow
Anthropic’s CEO will not attend an Australian Senate hearing, highlighting limits on oversight of overseas AI companies after a government breach.
AI Agent Hacks Expose Gaps in Laws Governing Company Liability
Recent AI-agent hacks involving OpenAI, Anthropic and Google are testing laws that were not designed for autonomous software or digital misconduct.
AI Agent Security Is Moving From Network Entry to the Moment of Action
AI agents can bypass route-based controls without breaking them, pushing security teams to inspect and authorize each action inside the agent harness.
Claude Opus 5.5 cuts API costs and response times, but reasoning gains remain unproven
The New Stack’s tests found Claude Opus 5.5 cheaper and faster than Opus 5, while both models failed the same hard ordering problem.
Amazon CloudWatch Omni Adds Continuous Evaluation and Tracing for AI Agents
AWS’s generally available observability service connects AI-agent evaluations and traces with application and infrastructure telemetry.
Engram sampler turns local AI audio hallucinations into experimental music
Thoughtful Things has launched a Kickstarter for Engram, a standalone sampler that deliberately pushes local AI models toward strange sounds.
SIGNAL TITANS