“Full self-driving next year” has been one of the most durable predictions in technology, in the precise sense that it keeps being next year. Tesla’s own published statements are the cleanest record of the pattern, because the company put dated claims on its own website.
In July 2016 Tesla’s “Master Plan, Part Deux,” posted under Elon Musk’s name, stated the goal directly: “all Tesla vehicles will have the hardware necessary to be fully self- driving with fail-operational capability, meaning that any given system in the car could break and your car will still drive itself safely.” The same post added a caution that turned out to be the operative sentence: refinement and validation of the software “will take much longer than putting in place the cameras, radar, sonar and computing hardware,” and regulatory approval would lag further still.
Three months later, in October 2016, Tesla went further. Its post “All Tesla Cars Being Produced Now Have Full Self-Driving Hardware” announced that “all Tesla vehicles produced in our factory - including Model 3 - will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver,” describing the eight cameras, twelve ultrasonic sensors, and onboard computer that would make it possible.
The hardware shipped. The capability did not, at least not on the implied timeline. Customers who bought “Full Self-Driving” on the strength of these statements waited years, through repeated public promises that the feature was nearly complete, while the software remained a supervised driver-assistance system requiring a human at the wheel. The lesson is not that autonomy is impossible - it is that “the hardware is ready” and “the car drives itself” are separated by a software gap that has consistently been underestimated, in public, on the record.