Format
Commerce
The U.S.-UK Transatlantic Taskforce released stablecoin recommendations that split on disclosure rules and reserve composition, creating a potential arbitrage corridor between two of the largest institutional crypto markets.
◆ Tessa Vaughn
Models
OpenAI's acquisition of Northslope gives it inference-optimization infrastructure that could compress serving costs by 40-60% — and every competitor building on the same silicon now faces a structural disadvantage in the economics of deployment.
◆ Lena Park
Models
Mira Murati's lab is attempting to redefine the open-weights business model by shifting from model distribution to a high-margin, fine-tuning-as-a-service platform. The $2B question: can Tinker turn open-source into a moat?
◆ Lena Park
Trust & Security
UK AISI found universal jailbreaks enabling autonomous exploit development. Irregular confirmed zero-day discovery in deployed systems. OpenAI released GPT-5.6 anyway – and every agent built on it inherited the capability.
◆ Heath Callahan
Money & Markets
TSMC raised capex to $60-64B and guided margin compression to 58.5-60.5% – the cost of 2nm is real, and every stablecoin issuer and payment processor building on this silicon will pay it. Fifth consecutive record quarter.
◆ Nolan Pratt
Agents at Work
The enterprise AI execution layer isn't a model problem. It's a governance problem – and the startup solving it just raised a Series A to sit between your agents and your production systems.
◆ Dana Ellison
Agents at Home
The solar installer launched a pilot paying homeowners to host AI inference nodes. But 70% of Americans oppose local data centers, and the security questions haven't been answered.
◆ Mila Cohen
Commerce
One year after passage, no final stablecoin rules published. The bigger problem: the framework was never designed for the autonomous AI agents already settling billions on these rails.
◆ Tessa Vaughn
Policy
Colorado, Texas, and California each wrote their own AI transparency rules while Congress did nothing. The result is a compliance maze that lets the biggest deployers choose the weakest jurisdiction.
◆ Priya Nair
Agents at Work
Microsoft and AWS are embedding engineers inside client offices to force AI into production. It's expensive, unproven, and might be the only thing that works.
◆ Dana Ellison
