Simon Willison Details How Coding Agents Are Ending Software Platform Lock-In

Simon Willison

May 15, 2026 · Updated Jun 12, 2026

Simon Willison highlights a shift where coding agents are used to perform full-scale architectural rewrites at a fraction of traditional costs. This capability makes high-stakes technical decisions reversible, as teams can now port codebases between platforms with minimal friction.

Simon Willison, the co-creator of Django, documented a case where a medium-sized technology company used coding agents to rewrite its legacy iPhone and Android apps to React Native. While the framework has improved, the breakthrough was the realization that the decision was no longer permanent.

Historically, choosing a framework created significant lock-in because migration costs were prohibitively high. As Willison's agentic engineering patterns become more reliable, the economic barrier to switching platforms is collapsing. This mirrors recent moves to facilitate OpenAI's agent workflow migrations between competing ecosystems.

This transition changes how teams evaluate technical debt and long-term commitments. If an architectural choice fails, agents can port the code back to its original state or a new platform. This flexibility allows developers to treat major stack changes as reversible experiments rather than high-risk, one-way doors.

View the full update on simonwillison.net

Simon Willison

@simonwMay 14

Mitchell's post here reminded me of a similar conversation I had recently about how cheap it can be to port native mobile apps to React Native using coding agents... and then port them back again later if it turns out not to work out https://t.co/p4xZ6bNqHi

9171

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

Keep reading

Simon Willison's Guide Argues Coding Agents Can Raise Code Quality

Shipping worse code with AI agents is a choice. Simon Willison's Agentic Engineering Patterns guide shows how teams can use agents to tackle refactoring debt — API redesigns, concept renames, module splits — that never felt worth the manual effort.

Guillermo Rauch reports coding agents are re-engaging enterprise leadership in development

Guillermo RauchMay 31

Guillermo Rauch reports coding agents are re-engaging enterprise leadership in development

Vercel CEO Guillermo Rauch reports that CEOs and CTOs are returning to active coding using agentic tools like Claude Code. This shift increases C-suite visibility into technical infrastructure, making software quality and stack effectiveness transparent to an organization's top leadership.

Andrew Ng argues agentic coding expands software jobs and shifts bottlenecks

Andrew NgApr 15

Andrew Ng argues agentic coding expands software jobs and shifts bottlenecks

Andrew Ng argues that agentic coding is expanding the software engineering field rather than shrinking it, citing rising job postings. This shift moves the primary development bottleneck from the act of writing code to the strategic decision of what to build.

OpenAI Team Shipped a Full Product Using Zero Human-Written Code

swyxFeb 11

OpenAI Team Shipped a Full Product Using Zero Human-Written Code

OpenAI's engineering team built a real internal product entirely with Codex agents — 1,500 pull requests, roughly a million lines of code, zero manual coding. Engineering shifted from writing code to designing agent environments and feedback loops.