4 ms·
A meta ride share app needs development effort (money), infrastructure (money) maintenance (money) and integration with local co-ops (money) and all that money
by frumiousirc 8y ago
A meta ride share app needs development effort (money), infrastructure (money) maintenance (money) and integration with local co-ops (money) and all that money needs an administrative hierarchy (more money).
All this overhead needs to be put in place without any direct income from riders. Either it come's in the form of a service that co-ops subscribe to and pay for or it comes in the form of a collective formed of individuals already in the co-op. Whatever the formation, it's activity that needs funding from the drivers.
This "meta" layer is then ready for abuse. Once the drivers are connected to this top-level coordinating layer the individual co-op layers become mere middle-men and are ripe for cutting out. Then we just get back to Uber.
It would be interesting to learn ideas how such a meta system could be created while limiting the overhead costs levied on the co-ops and retaining a decentralized cooperative system that allows drivers (and the higher layer workers) to retain their power.
- zanny 8y agoThere is absolutely no reason for ride sharing to not be a protocol. Drivers join a federated network of other drivers they "agree" with and you pick the driver pools you want (and they could be local drivers, or they could be drivers that meet some criteria or vetting process). You could have any number of ride share apps that use the protocol and interface with server infrastructure of each network. If you distribute the authority in the system you decentralize the power structures that produce hostile actors like the brass of Uber and Lyft. Nothing says each city can't have its own app and own driver cooper so long as my app from my city can speak your cities language when I'm there.