ai.
22 writings found
Latest Archives
SilverTorch: How Meta Rewrote the Rules of Recommendation Systems
Meta's radical shift from microservices to a unified neural network transforms retrieval at scale.
Vibe Coding: When We Stopped Reading Code and Started Trusting the Vibe
A deep dive into Andrej Karpathy's vibe coding concept - what it means, why it matters, and why forgetting code exists might be both its greatest strength and deepest flaw for developers.
Datasette Agent: When Three Years of Tooling Finally Converge
How the intersection of LLM Python libraries and Datasette creates a new paradigm for conversational data exploration.
ERA: How Google's AI is Reshaping Empirical Science and What It Means for Us
Google's Empirical Research Assistance uses AI to automate scientific coding. Here's what ERA really changes about how we do research and build intelligent systems.
Six Months of LLM Chaos: When Coding Agents Became Usable
November 2025 marked an inflection point in LLMs. Coding agents crossed a threshold from novelty to daily driver, while laptop-available models shattered expectations.
Beyond the Hype: How LLMs Are Reshaping Software Architecture
From legacy migrations to agentic patterns, exploring what agentic programming means for how we actually build systems and mentor the next generation.
Using LLMs as Interrogators: A Better Way to Extract Complex Knowledge
How to use AI to interview humans and build better context documents, turning conversation into structured knowledge that scales.
Open Science as Infrastructure: Why Google's Research Philosophy Matters to Developers
Exploring how open-source datasets and collaborative science are reshaping the developer ecosystem and enabling the next generation of breakthroughs.
Tests Are the Real Safety Net: Why Your AI Specs Need Executable Validation
Writing specs for LLMs is trendy. But without automated tests, you're flying blind. Here's why the spec document isn't your safety net.
Building macOS Apps Without Knowing Swift: What Vibe Coding Actually Teaches Us
I built two monitoring tools for my M5 MacBook using Claude and GPT without writing Swift myself. The results work, but should they?
Google's Vibe Coding XR: When AI Writes Your Spatial Computing Apps in 60 Seconds
Google Research launches XR Blocks with Gemini integration, letting developers prompt their way into physics-aware WebXR apps. I dig into what this means.
Starlette 1.0 and the Problem of Training Data Obsolescence
Starlette finally hits 1.0, but breaking changes expose a fascinating problem: how do you make LLMs generate code for frameworks they weren't trained on?
Starlette 1.0 and the Problem of Teaching AI New Tricks
Starlette finally hits 1.0, but breaking changes expose a fascinating challenge: how do you get LLMs to generate code for versions they weren't trained on?
Starlette 1.0 and the curious case of teaching AI new tricks
Starlette finally hits 1.0, but what happens when your LLM was trained on outdated code? Claude's new skills feature might just solve that problem.
Automating Security Fixes at Billions-of-Users Scale
How Meta's security team uses AI to patch vulnerabilities across millions of lines of mobile code without driving engineers insane.