💻 Assetnote engineer talks about how AI takes the joy out of development
Shubhankar (Shubs), an engineering manager at Assetnote (a company developing in the field of code security auditing), published an essay and a real dialogue with his subordinate about how work is changing in the age of AI. The engineer lists specific losses: previously, he fully understood every commit he sent, kept the architecture of the codebase in his head, and felt a sense of ownership over the result. Now, AI tools are "eating" critical thinking, their "memories" do not remember fixes, and models and tooling change every ~6 months, devaluing deep customization. The author himself, who has spent 15+ years developing the skill of source code auditing, admits that he is demoralized by how quickly AI has excelled in his field.
🌍 This is not another hype article about development speed, but a primary document — a real exchange between a manager and an engineer. The diagnosis is formulated directly: AI is turning engineers into "middle managers" — review and planning instead of design and debugging, which many do not want. For employers, this is a signal to reconsider the role of the engineer and the structure of teams, not just to implement tools. The caveat is important: this is a personal essay and one dialogue, without quantitative metrics, so generalizations are the author's interpretation, not a measurement.
👤 From the text, you can take techniques that can be applied immediately: instead of weak AI-memories, store instructions in CLAUDE.md / AGENTS.md files, which are forcibly included in every subsequent prompt; direct training toward AI skills, not code for the sake of code; increase the ambition of tasks — give AI problems "even harder than before," up to previously unsolved mathematical hypotheses and the search for zero-day vulnerabilities. The author calls deep pre-AI understanding of engineering "artisanal programming" and advises seeing it as a long-term competitive advantage.
Source 1: https://shubs.io/do-we-still-enjoy-software-engineering-in-the-age-of-ai/