I get the impression that Starship is doing test deployments in many different cities. The videos are probably indeed intended as press pushes, but they make a lot of them, they try a lot of different locations and conditions, and they interview generously and with a focus towards business value, which speaks to the underlying tech being sound in principle. I would be much more concerned if they were showing the same laboratory environments over and over or focused their videos on unsubstantial claims about the tech - they don't do either of those things. Here's some more recent coverage: https://www.youtube.com/watch?v=1GpaAF3_lKA
And while there are some obvious limitations to wheeled bots as a door-to-door courier system, the market for "personal luggage that follows you", which is more like what Gita is trying, could also be a pretty good one, and less ridiculous looking than "luggage you ride," [0] our current incumbent for luggage innovation. Having a bot stay by you has the gratifying feeling of having a pet or perhaps a servant.
These small scales and low speeds are much more amenable to experimenting, making mistakes, and ceding control than a highway-speed automation, as well. At worst, they're toys, but toys are often a good starting point for serious stuff too. It makes for very efficient R&D.
"Golf cart that follows you" is a thing.[1] Those have been around for 10 years, and they sell modestly. Suitcases which follow you exist, but haven't caught on.
It might be useful to have a set of suitcases that would follow the user and each other like a train. That would be an amusing option for road cases for rock groups and trade shows.
And while there are some obvious limitations to wheeled bots as a door-to-door courier system, the market for "personal luggage that follows you", which is more like what Gita is trying, could also be a pretty good one, and less ridiculous looking than "luggage you ride," [0] our current incumbent for luggage innovation. Having a bot stay by you has the gratifying feeling of having a pet or perhaps a servant.
These small scales and low speeds are much more amenable to experimenting, making mistakes, and ceding control than a highway-speed automation, as well. At worst, they're toys, but toys are often a good starting point for serious stuff too. It makes for very efficient R&D.
[0] https://www.youtube.com/watch?v=npEkbCmE31Y