That's true, IF passengers are interested in sharing a ride with two other passengers (and if they're interested in trading cost savings for the delays involved in that).

I kind of doubt that there will ever be a day when rides have an average of 3 passengers. Figure that if your base fare is $20, the first person you share with saves you $10 (or maybe $9 or something if you have to shore up the profits of the hailing app). Second person you share with saves you... $3. Third person saves you $2.

The returns to the passengers for additional passengers falls off fast, and the costs to the passengers (in terms of time, of discomfort, etc.) are linear or even more than linear. So probably not three passengers.

But two passengers is possible and would make for a real network effect... if ultimately there's enough demand for shared rides at all. There might be! Hard to say.

I was refering to three passengers sharing. Versus two passengers sharing it's a cost decrease of 33%. Maybe over one time trip it doesn't matter much, but if you're doing it alot ,for many 33% of their total transportation budget starts to matter.

But the end goal is to replace a large part of private transportation. And with car occupancy of 3 passenger and a drivers ,maybe cars start to get special road rights(as a tool to decrease traffic - it's part of some future roads in israel).

At that point, with UBER running a large part of the traffic and being able to create optimal routes, together with special road right , the difference in delay might not be that big versus driving , and you don't need to drive.

Yes, there are many if's here, as in all visions. but it might come to pass.