If All Software Is Being Rewritten, Sequence Matters

When a CEO of Microsoft's scale says 'all software is being rewritten,' it's worth taking seriously—not as hype, but as a signal about where engineering effort is actually flowing.

The premise is straightforward: AI capabilities are becoming native to development workflows. That means existing codebases, architectures, and even development practices are becoming candidates for redesign. Not because they're broken, but because the cost-benefit math has shifted.

For enterprises, this creates a specific challenge: You can't rewrite everything at once, and you shouldn't try to. The harder question is sequencing. Which systems benefit most from being reimagined with AI as a first-class input? Where does the effort actually return value in the next 12-18 months versus becoming a multi-year debt project?

I've seen teams get pulled in two directions—either treating this as a check-the-box modernization play, or getting paralyzed waiting for the 'right' moment to start. The middle ground is usually more productive: Pick one meaningful system, understand what changes when you build it differently, then apply those lessons systematically.

What's your organization actually rewriting right now, and what's driving that priority?

Source: The Motley Fool