It makes sense to target ridesharing while the tech is being developed because:

(a) you can amortize the large up-front cost of the hardware over many more trips per day.

(b) you can geographically restrict where the vehicle operates to areas you've mapped in detail and know to be relatively safe

(c) you can collect lots of raw data for training and allow remote operators to assist if the vehicle gets stuck (many people would have privacy concerns if their personal car was doing this).

Over time, the hardware cost will come down, geographic availability will increase, and the need for remote assistance will decrease. Then you might start to see ones you can fully own.

At that point, though, the question becomes would you want to own one? Particularly if ride-share vehicles are ubiquitous and you can nearly instantly summon one that's exactly the type you need no matter where you are.

Vehicles also act as large portable storage, which is very convenient when you have kids. e.g. stroller, diaper bag, change of clothes, sports equipment, emergency supplies. It would be annoying to always have to keep things on your person. With a personal vehicle, you also don't have to worry about something falling out of a bag and being lost forever after the vehicle drives away, or accidentally leaving a piece of equipment in it; just go check the parking lot. Or what if you want to make a trip to multiple stores and want to keep bags from the first in the car while you're in the second?

EVs also have the potential to act as large backup batteries for your home.

You can also choose how to amortize your costs. My mom's been driving the same car for 22 years now.

I could certainly see a scenario years down the road where having your own self driving car is a luxury/status symbol.