Tesla's was a lot of things. It was a shockingly deep dive into hardware architecture, it was a lesson on training a neural network, and most importantly, it was a look at Tesla's plans to have a massive fleet of "robotaxis" operating in the US.
Musk's roadmap to having his autonomous taxi fleet is surprisingly direct and well underway. The first big hurdle for any manufacturer attempting something on this scale would be hardware. Tesla gets around that by having installed its self-driving hardware in cars since 2016 -- it's the same hardware that makeswork so well.
The way that a Tesla robotaxi fleet would work, according to Musk, is that Tesla owners could opt into having their car become part of Tesla's taxi fleet when they weren't using it. The car would then drive off and do ride-hailing stuff while its owner sat at work or home.
Tesla would charge users for their rides, obviously, and then would take a cut of the car's earnings off the top. The rest would go to the vehicle owner. In areas where there weren't enough vehicle owners willing to let their vehicles join the fleet, Tesla would have its own self-managed fleet in place, likely made up of off-lease cars that it's no longer letting customers buy.
All of this would be controlled through the Tesla app, and we get the feeling that, apart from the lack of a driver -- and a driver's penchant for too many air fresheners and awkward conversation -- it would feel like a standard ride-hailing service.
Musk plans to have Tesla's full self-driving capability "feature complete" by the end of 2019, with it being ready for unsupervised use by Q2 of 2020 and -- this is the big one -- 1 million robotaxis on the road testing without passengers by the end of 2020, pending regulatory approval.
So what will the economics of this all look like when Musk flips the robotaxi switch? According to his presentation, the yearly income from an autonomous taxi would be around $30,000 gross. Even if you take into account things like wear and tear, and the usage of consumables (which,, isn't much), that's still a tidy little sum per annum.
So, what do you think -- will we soon see Tesla's robotaxis zipping around our city streets? Or is this a case of Musk mucking up his timeline, and robotaxis are still a long way off? Let us know in the comments section.