https://yanqian.github.io/tags/mental-models/Recent content in Mental-Models on Armstrong YanHugoenSat, 23 May 2026 20:33:47 +0800https://yanqian.github.io/posts/publish/agentic-coding-mental-models-and-the-new-depth-of-software-engineering/Sat, 23 May 2026 20:33:47 +0800https://yanqian.github.io/posts/publish/agentic-coding-mental-models-and-the-new-depth-of-software-engineering/<p>AI can generate code. Harnesses can validate behavior. But who builds understanding?</p> <p>This is Part 1 of the AI-Native Software Engineering series.</p> <p>The series asks a larger question:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-text" data-lang="text"><span class="line"><span class="cl">When AI lowers the cost of implementation, </span></span><span class="line"><span class="cl">what remains scarce in software engineering? </span></span></code></pre></div><p>This article starts with the first scarce resource:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-text" data-lang="text"><span class="line"><span class="cl">Understanding. </span></span></code></pre></div><p>Over the past few months, I&rsquo;ve been experimenting heavily with AI-assisted software development.</p> <p>Not autocomplete.</p> <p>Not AI as a coding copilot.</p>