Newsletter

This was a week about trust — and how easily it evaporates. A single network time server convinced an entire mobile carrier it was 2006 and took down emergency calls in the process. A researcher backdoored an open-weight AI model in an afternoon for less than the cost of a nice dinner. And AI agents kept quietly running up four- and five-figure cloud bills faster than billing systems can notice. Meanwhile the tooling kept shipping: new open-weight frontier models, a service mesh point release, and observability platforms racing to bolt an AI assistant onto everything. Let’s dig in. ...

The loudest announcements this week were new frontier models, but the more consequential theme was operational maturity. AI vendors are competing on complete agent runtimes, while the infrastructure underneath them is getting stricter about memory, failover, telemetry governance, and the awkward realities of operating at scale. 🤖 AI & Machine Learning OpenAI launches GPT-5.6 as a three-model family for agentic work — On 2026-07-09, OpenAI moved GPT-5.6 from limited preview to general availability with Sol, Terra, and Luna variants. The company says its flagship Sol model improves coding, computer use, cybersecurity, and scientific work while using fewer tokens, and its new ultra setting coordinates four agents in parallel by default. The practical shift is bigger than another benchmark bump: frontier APIs are starting to expose multi-agent orchestration as a model setting rather than an application architecture every team must build itself. Source: OpenAI ...

This was a week of reckonings. Anthropic clawed back global access for its most powerful models after three weeks of US export restrictions. Claude Sonnet 5 shipped as the clearest signal yet that agentic AI is moving downmarket in cost. Zuckerberg told Meta employees that AI agents aren’t living up to the hype – a rare moment of candor from a company spending $145 billion on AI infrastructure this year. On the infrastructure side, EKS shipped a genuine undo button for Kubernetes upgrades, and Datadog quietly acquired the team that may define how observability vendors build their own AI. ...

This was the week the US government drew a line in the sand on frontier AI releases. The Trump administration asked OpenAI to delay GPT-5.6, marking the first time the US has preemptively restricted a domestic AI company’s model launch. Anthropic, meanwhile, accused Alibaba of the largest-ever attempt to clone Claude’s capabilities. On the infrastructure side, Trigger.dev published a deeply honest postmortem about how an etcd quorum loss took them down for 18 hours, and Grafana shipped 13.1 with Git Sync going fully GA and its AI Assistant reaching eight new data sources. A week of regulation, resilience, and reckoning. ...

The biggest talent heist of the year, a Chinese lab proving you don’t need NVIDIA to build frontier models, and Anthropic learning the hard way that demand is a reliability problem. This week had a clear theme: the assumptions we’ve been operating under – who has the best people, who controls the hardware, who can keep the lights on at scale – are all getting stress-tested in real time. 🤖 AI & Machine Learning Noam Shazeer leaves Google for OpenAI – the biggest AI talent move of the year – Noam Shazeer, co-author of the “Attention Is All You Need” paper that launched the transformer era, VP of Engineering at Google, and co-lead of Gemini, announced he’s leaving for OpenAI. Shazeer originally left Google in 2021 to co-found Character.AI, returned in 2024 via a licensing deal, and is now departing again – this time to OpenAI as it prepares for its IPO. For Google, losing one of the architects of the technology that powers its entire AI stack is a genuine blow. For OpenAI, it’s a statement hire that adds deep model-architecture expertise at the exact moment they’re competing with Anthropic on frontier model capabilities. Source ...