Industry · Software · Tech Policy · Events
Engineering culture, developer tooling and the working lives of the people who ship software, from The Silicon Ledger.
Perplexity's Portable Computer, an on-device AI agent for complex tasks, is now available on Windows systems equipped with Nvidia RTX GPUs carrying at least 24GB of memory.
Claude Fable 5.1 wins a new coding benchmark with a 38.8% success rate, but that means it fails nearly two-thirds of the time when tested on proprietary code from real companies.
GitClear's analysis of 623 million code changes reveals that AI-assisted development boosts output modestly while triggering a dangerous shift toward copy-paste coding and away from refactoring, threatening long-term maintainability.
As Chip Huyen returns to P99 CONF, her framework for reducing AI inference costs without hardware upgrades remains relevant—though the rise of agents and reasoning models has shifted how practitioners think about the problem.
An AI system can return a successful response, pass all evaluation metrics, and still deliver incorrect information to customers. Understanding why requires looking beyond code changes to the execution trace itself.
As Model Context Protocol spreads across enterprise infrastructure, security experts are realizing that the vulnerability isn't in the protocol itself—it's in how organizations grant access to AI agents.
OpenAI has recruited the creators of Git AI, a tool that quantifies code output from AI agents, to help enterprises understand the business value of its Codex product.
Anthropic's AI-Native SDLC Playbook correctly identifies that code generation is no longer the constraint in software development, but organizations need adaptable approval processes tailored to different types of changes rather than rigid, universal workflows.
