Open Driver
Current state of driving with uber/lyft: https://www.youtube.com/watch?v=OEXJmNj6SPk
Concept
The system would allow riders to broadcast ride requests while keeping their specific location private. Information such as the total distance, major roads involved, and other relevant metrics (like the percentage of time spent on main roads) would be exposed. This would allow for enough transparency for drivers to bid on requests while maintaining rider privacy until the driver and rider have made an agreement.
This could use ZK, MPC, FHE and/or some classic cryptography depending on implementation.
Driver On-boarding
Drivers could be directly on-boarded using TLSNotary from Uber, Lyft, Grab, or any other platform.
Driver Bidding System
Drivers can bid on ride requests. The bids could be made manually or through automated bidding algorithms. Algorithms might take into account factors such as location, distance, fuel costs, and personal preferences. Developers could compete to offer optimal algorithms tailored to specific areas or vehicle types. Transparency of algorithm performance and fairness could be enforced through cryptographic commitments to ensure developers are running the algorithms they claim to be.
Rating and Incentives
Drivers and customers would rate each other, and drivers could be rewarded based on their ratings and performance. A fee system could be implemented that charges a small amount per ride and those fees could be used to incentivize better drivers, giving them stronger bidding power. The platform’s fees should be transparent, where infrastructure costs could be proven and paid for via fees(using methods like TLSNotary), and excess revenue redistributed to top-rated drivers rather than middlemen.
Open Ecosystem
The platform should allow anyone to offer frontends and bidding algorithms (potentially for a fee), enabling competition that would cater to different languages, regions, or use cases. Marketplaces for algorithms and services could emerge without fracturing the number of available drivers or
thought dump below, please excuse the rambling, but wanted to keep for those interested
Make a system where people can openly request a ride and have their inputs be private or masked as to not share their address,but expose the distance and what major roads they will be on.
Then drivers can bid for that ride, where the ride can then later be privately reviewed.
There will be a fee system that would pay good drivers more out of that fee,giving them more bidding power.
Any algorithms used can be committed to via zk. There could be a market of algorithms for automatic bidding algorithms that take into account your location,their location,the price of gas, etc
FHE, MPC,and zk could be used in a lot of places here, so could ezkl for the bidding algorithms.
Any commitments and registrations can happen on chain.
We could push mobile proving, we could demonstrate all this cryptography, and get rid of the rent seekers.
Normally I am pretty skeptical of thinking an app would have quick adoption, but Uber and Lyft are screwing drivers and I think any system that was provably more fair would get quick adoption.
This is another idea that I have that would really fit in here. We could use TLSNotary to prove the AWS bill to run the infrastructure and the fees collected could help reimburse the cost to run the infrastructure but no more,all additional fees would go towards the good drivers. We could make it so let's say last months bill was $200, we would split that into 672 (24 hours x 28 days) increments, $0.30 per hour. That money could be taken out of the fees collected per hour and the rest could go towards subsiding good drivers based on their rating.
All drivers could "go on the clock" and "go off the clock" just like on Uber and Lyft, say they would be willing to service some area, and their rating would go into a pool of other drivers that are also on the clock when you propose a ride.
Private set intersection could be used to compare the area someone is being picked up in and the service area of a driver.
The algorithm to search for a ride could reveal information like the major roads, the distance, how many roads, and the % of time on the most major road. This would hopefully give enough information to be pretty private, but also give automatic bidding algorithms enough data to help maximize drivers profits. Instead of the algorithms making Uber and Lyft rich, it would give the drivers a decent shot at making a living.
From the customers perspective, they would broadcast their ride in zk (mentioned above) and drivers bids would begin to show up, all drivers bids could contain a timeout and be cancelled if they pick up another ride. The customer would select a bid and accept it and the driver would have to acknowledge it for the ride to be accepted.
At first the drivers could manually enter bids,but we absolutely would need algorithm support. Ezkl or zk stats could be used to prove that people are actually executing the algorithm they say they are paying for, and not running one algorithm for the first month of a subscription and degrading it later. All algorithms should be committed to publicly. Profits, algorithm fees, and reviews of algorithms should also be completely public.
Algorithm devs could compete to offer the best algorithm for a certain area of a country or city, or with a certain type of vehicle, or particular day of the week (best for surge pricing Saturday night).
Identities of the driver and the passenger could be split between the service, the driver, and the passenger, so that if the driver or the passenger needs to reveal the others identity for any legal or safety reasons, we could.
Anyone could offer a service, like an app, for a fee per ride or per month, and front ends/apps could compete for best service in their area/language.
The biggest issue I see is attracting customers, but if you cut out the middle man their prices should be cheaper.