We should be more ambitious with the software we build

Michael Malis's pgrust project demonstrates that ambitious software rewrites are worthwhile despite widespread skepticism about AI-generated and low-quality projects, as evidenced by the substantial bugs it has uncovered in PostgreSQL itself. While most rewrite projects are poorly executed, Malis's thorough testing approach shows that ambitious builders can produce genuinely useful software if they commit to rigorous testing and benchmarking, potentially improving existing systems even if the project itself doesn't fully succeed.