Former google engineer details how AI tools wrote over 30,000 lines of code for him

Wang Yonggang, a former senior engineer at Google, detailed in his article Farewell, Programmers how artificial intelligence is fundamentally changing the role of programmers. When an American computer teacher requested improvements to an Altair 8800 simulator he had open-sourced six years prior, Yonggang tasked Claude Opus 5.5 with the system design. The AI not only completed the initial work but also suggested adding features like an assembler and a debugger, resulting in a project exceeding 30,000 lines of code, complete with unit and browser tests, with Yonggang only needing to confirm the approach and provide feedback. Furthermore, Yonggang noted the AI's capacity for independent research, citing an instance where it helped reconstruct Leibniz's mechanical calculator by contacting a German museum and acquiring 135 pages of historical scans. Yonggang believes that as AI programming advances, the value of traditional technical experience is being re-evaluated. He predicts that future development teams may become small groups of product and business leads managing numerous AI agents, suggesting the career path for conventional programmers faces significant contraction.