Commentary

For most of the history of software engineering, we’ve built systems around a simple and comforting assumption: Given the same input, a program will produce the same output. When something went wrong, it was usually because of a bug, a misconfiguration, or a dependency that wasn’t behaving as advertised. Our tools, testing strategies, and even […]

Over the past six months, I’ve watched the same pattern repeat across enterprise AI teams. A2A and ACP light up the room during architecture reviews—the protocols are elegant, the demos impressive. Three weeks into production, someone asks: “Wait, which agent authorized that $50,000 vendor payment at 2 am?“ The excitement shifts to concern. Here’s the

This post first appeared on Addy Osmani’s Elevate Substack newsletter and is being republished here with the author’s permission. AI coding assistants have quickly moved from novelty to necessity, where up to 90% of software engineers use some kind of AI for coding. But a new paradigm is emerging in software development—one where engineers leverage

At a private dinner a few months ago, Jensen Huang apparently said what I’ve been thinking for some time. The US is significantly behind China in AI development. Here are some of the reasons. Huang starts with the ratio of AI developers in China (he estimates 1 million) to AI developers in the US (20,000).

Papers on agentic and multi-agent systems (MAS) skyrocketed from 820 in 2024 to over 2,500 in 2025. This surge suggests that MAS are now a primary focus for the world’s top research labs and universities. Yet there is a disconnect: While research is booming, these systems still frequently fail when they hit production. Most teams

Papers on agentic and multi-agent systems (MAS) skyrocketed from 820 in 2024 to over 2,500 in 2025. This surge suggests that MAS are now a primary focus for the world’s top research labs and universities. Yet there is a disconnect: While research is booming, these systems still frequently fail when they hit production. Most teams

This post first appeared on Nick Tune’s Medium page and is being republished here with the author’s permission. Example architecture flow reverse-engineered by Claude Code I have been using Claude Code for a variety of purposes, and one thing I’ve realized is that the more it understands about the functionality of the system (the domain,

The hard truth about AI scaling is that for most organizations, it isn’t happening. Despite billions in investment, a 2025 report from the MIT NANDA initiative reveals that 95% of enterprise generative AI pilots fail to deliver measurable business impact. This isn’t a technology problem; it’s an organizational design problem. The reason for this systemic

For 30 years, digital commerce has been a destination. We “go to” a website, a marketplace, or an app. In this single, bundled environment, we handle discovery, comparison, and checkout. The entire architecture of the web, from product pages to payment gateways, is built on this assumption. This assumption is now facing its first real

This post first appeared on Mike Amundsen’s Signals from Our Futures Past newsletter and is being republished here with the author’s permission. We’re long past the novelty phase of AI-assisted coding. The new challenge is measurement. How do we know whether all this augmentation—Copilot, Cursor, Goose, Gemini—is actually making us better at what matters? The