Back in the Bay Area for the third straight week. I’m looking forward to seeing a bunch of colleagues tomorrow that I rarely see in person.
[blog] The New Security Agenda. a16z’s survey of 100 CISOs shows that AI-driven threats and new attack surfaces are driving a huge increase in cybersecurity budgets.
[blog] Making agents fast – what I’ve learnt over the last 12 months. Smart post. It looks at techniques around inference, tool calls, and even hardware.
[blog] Build Generative UI Apps with A2UI & CopilotKit Agents. Work with generative UIs regardless of agent framework.
[article] Stop Blaming Code Review for AI’s Delivery Problem. Yes, validation is the problem now, but it’s also about rethinking testing. It’s can’t just be your CI/CD pipeline where tests run.
[blog] Roblox’s approach to tripling AI code review acceptance rates. Seems like a good listen, and these recap points are useful.
[blog] First look at Cloud Run instances. Fantastic deep dive from Wietse. Get your own microVM that’s suitable for all sorts of workloads.
[blog] Work in progress is the new software delivery bottleneck. It’s tempting now to have a ton of WIP with agents, going against years of conventional wisdom. But we’re not necessarily getting better at validating all that parallel work.
[article] Google Cloud Growing 3.5X Faster than Microsoft Cloud as Customers Pick AI Leader. Lots of growth all around. But we’re growing more in new dollars and percentage rate than anyone else.
[blog] Default to serverless for AI agents, then let production data decide what moves. Makes sense. Figure out if you need a durable host, or can get by with straight-up serverless or abstractions like Agent Substrate.
[article] Go concurrency distilled. Great content if you want to really understand how Go handles concurrency scenarios.
[blog] How to Trust a Coding Agent: Read the Logs, Not the Summary. Probably good advice. The logs show the paths taken, which may not always match the conclusion.
[blog] Grounding agents in one context graph connecting semantic models you already trust. I really like this new functionality. It’s a powerful way to define a universal context engine.
[article] GitLab Bets Context, Not Bigger Models, Will Fix AI Agent Costs. Both? You need great context, but models with smarter reasoning will also eliminate a bunch of things we ask of our harnesses today.
Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:
Leave a Reply