The Visual Studio Skeptic

Twenty years ago, I was working on Visual Studio at Microsoft when I had an encounter that still sticks with me today. I was manning a booth at a trade show, demonstrating new IDE features in Visual Studio 2005, when a developer approached me with an unusual complaint.

‘Visual Studio has gotten too good at code generation,’ he said, dead serious. ‘You need to cut it out because building data access objects is what I do.’

I thought he was joking. Here was someone essentially arguing that Microsoft should make its tools less capable so he could keep doing boring, repetitive work. When I suggested that these automation features would free him up to focus on more interesting, higher-value problems, he wasn’t having it.

I think about that developer often these days, especially as we’re witnessing what might be the most significant shift in software development since IntelliSense was first added to Visual Studio. The parallels are striking: just as that skeptical developer worried that automated data access would eliminate his job, today’s developers express similar concerns about AI coding assistants.

The argument then was familiar: “If the computer can generate my data access code, what value do I bring?” The argument now sounds eerily similar: “If Claude or Copilot can write my functions, what’s my role?”

But here’s what that developer missed, and what some of today’s skeptics are missing: better tools don’t eliminate developers—they elevate what it means to be one. IntelliSense didn’t put programmers out of work; it made us faster and helped us write better code with fewer typos. Database wizards didn’t eliminate the need for developers; they freed us from tedious boilerplate so we could focus on business logic and user experience.

Today’s AI coding tools follow the same pattern. Claude Code doesn’t replace thinking about architecture, user needs, or system design. It handles the mechanical translation of ideas into syntax, the grunt work of scaffolding and refactoring, the tedious debugging sessions that used to eat entire afternoons.

The real question isn’t whether these tools will change how we work—they already are. The question is whether we’ll embrace them as force multipliers for human creativity and problem-solving, or resist them out of fear that automation somehow diminishes our craft.

I wonder if that Visual Studio skeptic from twenty years ago is still programming. More importantly, I wonder what he thinks about the fact that modern development—aided by AI—can accomplish in hours what used to take days or weeks. My guess is that if he stuck with it, he eventually discovered that better tools didn’t make him obsolete. They just made room for him to solve bigger, more interesting problems.