Waymo is not hype, and it is also not finished. Rides went tenfold in under two years, 50,000 to 500,000 a week, at a $126 billion valuation, and it added four cities in a single day. I was wrong to write self-driving cars off as vaporware in 2016.
But scaling a fleet and finishing a technology are different jobs. The 1-million-rides-a-week target is a stretch (an independent forecaster models 775,000), and 2026 shows the seams: three software recalls in a year, school-bus and pedestrian probes, and a dozen robotaxis stranded with dead batteries during San Francisco's Fourth of July fireworks. It works often enough to be useful and breaks in public often enough that you should read the recall notices before calling it done.
