AI-Assisted Engineering: My 2025 Substack Recap

Over the past year, I’ve written a handful of deep dives into the intersection of AI and software engineering on my Substack newsletter, Elevate. This post recaps some of the most-read pieces, in case you missed them or want a refresher. Each explores a different angle on how AI is changing the way we build software.

The 70% Problem: Hard Truths About AI-Assisted Coding

This article explores a common pattern: AI gets you 70% of the way to a working solution, but that last 30% is where things get tricky. I dig into the human-AI handoff, the limitations of current tools, and what it means for engineering velocity.

Beyond the 70%: Maximizing the Human 30% of AI-Assisted Coding

A follow-up to the 70% problem, this piece focuses on how developers can close the gap. I cover techniques for navigating ambiguity, debugging AI-generated code, and staying in the loop when machines do most of the typing.

MCP: What It Is and Why It Matters

My most recent post introduces the Model Context Protocol (MCP) - a conceptual “plug adapter” between AI models and different runtimes or environments. I explore how MCP can help standardize AI model interactions, making agent workflows more composable and less brittle.

Why I Use Cline for AI Engineering

This article walks through why I’ve been using Cline - a free VSCode plugin - for serious AI engineering work. I cover what sets it apart, the pros and cons, and why tooling that understands both your code and your intentions is critical.

Future-proofing your Software Engineering career

This popular article provides practical strategies for engineering careers in an AI-driven future. I explore how software engineers can adapt their skills, mindsets, and approaches to stay relevant and thrive in a rapidly evolving technical landscape where AI is changing the nature of development work itself.

Leading Effective Engineering Teams in the Age of GenAI

This pragmatic guide for Software Engineering leaders explores how to navigate team leadership in an era where AI tools are transforming development workflows. I discuss strategies for integrating AI tools effectively, managing the changing dynamics of technical work, and ensuring teams remain productive and engaged as AI reshapes our engineering practices.

AI-Driven Prototyping: v0, Bolt, and Lovable Compared

This article examines three notable AI-assisted development tools - v0 by Vercel, Bolt by StackBlitz, and Lovable. I discuss their practical applications for bootstrapping MVPs, their limitations, and trade-offs from an engineering perspective. Each tool brings a unique approach to reducing the friction between ideation and implementation, catering to different needs in the development lifecycle.

Personal Software: The Unbundling of the Programmer?

This piece zooms out to ask what happens when AI lets anyone build software. I explore the idea that we’re not just empowering developers - we’re changing who gets to build. It’s about tools that adapt to people, not the other way around.

If you’ve read and shared any of these this year - thank you. I’m always experimenting with what AI means for developers and how we can make our tools more thoughtful, flexible, and empowering. More to come soon.