Makes sense to me and an interesting inversion of the problem. The semi industry is the best at engineering high throughput, reliable systems at scale. Maybe to do real automation we need to start by shrinking the problem first. Very cool work.
