“Anyone Can Code Now” – Netlify CEO Talks AI Agents

Netlify’s daily signups have skyrocketed from 3,000 to 16,000, and the twist is that most of these new users aren’t traditional developers. They’re marketers, designers, and product managers—people who, just months ago, couldn’t write a line of code. This explosion signals a fundamental shift: the barrier between having an idea and shipping working software is collapsing, not because everyone is learning to code, but because AI agents are becoming the new collaborators.

At the heart of this shift is what Netlify CEO Matt Biilmann calls “Agent Experience” (AX)—designing products and platforms for both human users and the AI agents that now routinely interact with them. For developer tools, this is becoming essential; their addressable audience has ballooned from 17 million professional JavaScript developers to the billions of people who use spreadsheets. The core skill of a developer is evolving from writing syntax to systems thinking, clarity of thought, and guiding AI to solve problems. Code itself is becoming less important, similar to how writing is a universal skill while professional writers remain a specialized subset.

This democratization is leading to a renaissance of creativity on the web. Contrary to the “dead web” theory—which predicts a landscape of only AI-generated and AI-consumed content—there’s a surge of people building “weirder, wilder, and more wonderful” projects, like intricate WebGL games, that were previously too complex to attempt. The browser is evolving into a true “user agent,” with AI acting as a personalized concierge for consuming the web. This suggests a future where software development is a common skill integrated into many jobs, while expert developers focus on higher-level architecture and design.

Surprising Insights

• CEOs are submitting pull requests again. Founders and executives with technical backgrounds are using AI coding agents to actively build features and products in between meetings, re-engaging directly with development in a way that was previously impossible due to framework complexity.
• The “dead web” theory appears backwards. While there is AI-generated content, the lower barrier to creation has led to a net increase in human creativity, with more unique, visually rich, and interactive websites and web apps being built and visited than in years.
• Users form strong preferences for their AI “concierge.” People are developing affinities for specific AI agents (like ChatGPT or Claude) and will prefer to interact with all services through that trusted interface rather than using a company’s custom AI, reshaping how consumption layers will work.
• Pricing models are scrambling. The industry is moving from recurring subscriptions to usage-based (token-based) models for AI tools, but there’s a concurrent search for “outcome-based” pricing, as token cost doesn’t neatly align with the value delivered to the user.

Practical Takeaways

• Focus on “Agent Experience” (AX). If you build a product, assume AI agents are now a core user persona. Optimize your documentation, APIs, and onboarding for both humans and autonomous agents.
• Cultivate systems thinking over syntax memorization. The new developer superpower is the ability to clearly define problems, articulate desired outcomes, and guide AI through iterative refinement—not recalling framework specifics.
• Start by remixing. Beginners should use tools like Bold or Builder.io to start “remixing” existing website patterns they admire, using AI to combine and modify them, which builds an intuitive understanding of structure and design.
• Interrogate the AI to learn. Don’t just accept the AI’s output. Ask “why did you do it this way?” and “how does this work?” Using AI with curiosity is a powerful way to accelerate your learning journey in software development.
• Embrace the frustration. The path to proficiency with AI development tools still requires a high tolerance for iteration and failure. The tenacity to keep refining prompts and approaches when things break is the enduring mark of a good developer.