Why every developer should learn to code and use AI

Executive overview

Most people frame AI as a reason to stop learning to code. The opposite is true. Coding is a foundational skill like literacy — knowing it lets you create, not just consume. AI makes coding more accessible and developers far more productive, but it doesn't remove the need for engineers who can reason about systems.

AI amplifies developers; it doesn't replace the need to understand how software is built.

Thomas Dohmke's path to GitHub CEO

  • Built his first app after Steve Jobs announced the iPhone SDK in 2008; quit a stable job at Bosch during the financial crisis to freelance
  • Co-founded Hockey App — a beta distribution and crash-reporting platform born from their own pain as mobile developers
  • Microsoft acquired both the Hockey App product company and the freelance contracting business in 2014, valuing the iOS/Android engineers
  • Moved from Stuttgart to Seattle in early 2015 with his family; seven of the original 11 employees eventually landed at GitHub

Why coding must be taught in schools

  • Software dominates daily life — travel, communication, waking up all depend on it
  • Being in "read-only mode" on technology is no longer sufficient for full participation
  • Learning to code doesn't mean becoming a developer, just as learning physics doesn't mean becoming a physicist
  • AI removes the English-language barrier to coding, democratising access globally

How AI is changing software development

  • GitHub Copilot began in June 2020 after GPT-3 was released; GitHub's focus is keeping developers productive and happy
  • The core developer frustration: a weekend idea expands into a months-long project because complexity compounds at every step
  • AI compresses that gap — very small teams (sometimes one developer) now believe they can build million-dollar businesses using AI agents
  • Most codebases carry two backlogs: new features and technical debt; AI helps work through both faster
  • Productivity gains of 10–50% are already visible in practice

Where AI still falls short

  • A single prompt cannot build a complex system like GitHub — thousands of architectural decisions are required
  • Choosing languages, frameworks, cloud providers, monolith vs. microservices: these still require human engineering judgment
  • Product-market fit, user experience, and profitability cannot be delegated to an agent today

Three things every developer (or aspiring developer) should do

  1. Learn to code — treat it as a fundamental skill, not optional
  2. Use AI to accelerate that learning; it has infinite patience and no judgment
  3. Keep training continuously — the field moves fast; staying current is as important now as it was in the 1990s

GitHub's scale and remote culture

  • 150 million users generate constant, high-volume feedback; filtering signal from noise is a core leadership challenge
  • GitHub has been remote-first since before COVID — a deliberate choice, not a pandemic response
  • Async collaboration via Slack, open-source norms, and video calls replaces email-heavy workflows
  • Being in Seoul while running the company is routine — location is largely irrelevant to day-to-day operations

More like this — when you're ready for early access.

Join the waitlist for a personal account and content recommendations based on what you're working on.

No spam. Unsubscribe at any time.

You're on the list. We'll be in touch before launch.

Get early access to the full library.

Join the waitlist for a personal account and content recommendations based on what you're working on.

No spam. Unsubscribe at any time.

You're on the list. We'll be in touch before launch.

Be among the first to get personalised recommendations tailored to your stage in business.

No spam.

You're on the list. We'll be in touch before launch.

Be among the first to get personalised recommendations tailored to your stage in business.

No spam.

You're on the list. We'll be in touch before launch.