The Human-First Engineering Manifesto
Writing code has never been the whole of engineering, and engineers have never worked without assistance. Compilers, IDEs, autocomplete, static analysis — each wave of tooling automated something that used to be manual. AI is the next step on that same path.
The fundamental question AI-assisted development poses is not new: what does it mean to be a great engineer when the tooling around us gets more powerful? The answer has always been the same. An engineer's value moves further up the stack — towards understanding, judgement, and the kind of contextual reasoning that tools cannot replicate.
📜 What we believe
🌱 We use AI to grow engineers, not replace them
Every generation produces great engineers. AI must not be the first that quietly fails to.
⚒️ We remember that writing code has never been the whole job
Engineering is design, reasoning, debugging, architecture, and judgement — not typing.
🧠 We think before we generate
Understanding comes first. AI accelerates execution; it does not replace the reasoning that should precede it.
🎯 We use AI with skill, not by reflex
Using AI well is a craft of its own. Model choice, context, and prompting all shape the result. Chucking a vague prompt into a chat box is not engineering. And the skill is not only in the prompt: a context window is finite and forgets, so preserving the understanding behind the work — as sessions end and conversations compact — matters as much as phrasing the request.
🪪 We own what we ship
Only humans can be accountable for quality, clarity, and safety. Every change an engineer ships has a named human owner, however it was generated. If you cannot explain it, you do not ship it.
🔍 We trust AI, but we verify everything
AI is confident, not always correct. Blanket trust is negligence. Blanket distrust is waste. The skill is knowing how much to trust the output.
📈 We grow through AI, not around it
AI should make us better engineers, faster — not let us coast behind it. A growth mindset is what turns the tool into mastery rather than avoidance.
🚦 We move fast without breaking clarity
Velocity matters only when it produces maintainable, coherent systems that the humans who come after us can understand.
💡 The central idea
AI amplifies engineers. It does not replace engineering. And engineering is a craft that must be grown — not just deployed.
Velocity without understanding is not progress — it is debt. An engineer who can direct an AI agent well but cannot reason about what it produces is not a senior engineer in waiting. They are a risk.
The engineers we fail to grow today are the senior engineers missing from the industry in a decade.
🌱 A commitment to early-career engineers
These beliefs carry an obligation to the people most exposed to this shift.
Early-career engineers are entering the industry at a moment when the tools available to them are extraordinary — and capable of masking gaps in their understanding in ways that catch up later. Good engineers don't happen by accident. The conditions that grow them have to be built and protected.
We commit to providing early-career engineers with:
- Regular access to people who will explain the why behind the what.
- Safe spaces to be wrong and to not know.
- Honest feedback that helps them grow, not feedback that protects their confidence.
- The recognition that struggle is not failure — it is the work.
The senior engineers of five years from now are the junior engineers of today. Protecting that pipeline is the reason this framework exists.
➡️ Next
Read the framework for how these beliefs translate into day-to-day behaviour.