The analysis is good. There is a reason Tesla is only at self driving level 2. While Waymo is at level 4. Even Mercedes Benz is at level 3. Tesla was supposed to pass Waymo in the self driving dust, even though Waymo had a 5 year head start. Tesla has over a million cars to Waymo's less than 1,000 cars on the road. That's a thousand times more data for its neural networks. But in Musk's infinite wisdom he removed radar and didn't use lidar. He said people don't use them so I don't need them. But people have a more developed neural network. He also wanted to cut costs, even though historically electronic device's costs have declined.
Intriguing analysis, but how is the situation affected by the fact that many rideshare drivers drive for both Uber and Lyft? And of course, most customers use both Uber and Lyft, and would add Waymo, when it becomes more widely available. Are there data on how many unique drivers drive for Uber and Lyft, and how many split their time between the two? In other words, is there double-counting in the estimates of the capacity of each platform?Likewise, how many riders are uniquely loyal to Uber or Lyft, and how many switch back and forth? And how big of a deal are the "frictions" that you mention (and which seem to annoy you with Uber and Lyft) to most customers -- Has anyone studied this issue? If these frictions are indeed significant to a large proportion of riders, then Waymo would offer a significant point of differentiation.
Once Waymo’s AV’s are on Uber, and assuming price benefits continue to flow to customers, Uber will have a steeper internal competition between drivers and AVs. With driver earnings already a fraction of total ride cost, it would be interesting to see how the platform continues to stay lucrative for drivers.
There is a research paper on this topic as well. At least initially, they will have to pay the remaining drivers more to keep them on the platform (since drivers will know that within a few years they will be displaced).
The evolution of urban mobility will indeed be a fascinating business case playing out in plain sight over the next decade(s) to come. In addition to the dynamics you mention, I would add a few other considerations.
1. I don't see Tesla being a player for a meaningful future (say 10 years). for both technological and regulatory reasons. Let's not forget that Tesla's FSD is mired at Level 2 AV operation (and still has a high accident/fatality rate), while Waymo is operating safely at scale at Level 4 (and about to add highway operation capability). Yes, I know Mr. Musk has previously pulled innumerable rabbits out of his hat before. Still, the technological challenge of operating Level 4/5 AVs in dense cities like NY, Sao Paolo and Mumbai is a LOT harder than catching a reusable rocket on a launch gantry. The former has to cope with the extreme volatility of human behavior and unpredictably variable micro-level edge-case circumstances while the latter is "just" physics in a controlled operating environment.
On the regulatory front, we haven't even begun to sort through the insurance, risk management and operational implications of allowing individual vehicle owners the freedom and ability to determine urban mobility supply on a totally decentralized basis. Fortunately, the regulatory concerns are moot for the foreseeable future, given Tesla's technological challenges.
2. For now, both Uber and Waymo have unique competitive advantages that are valuable to each other. Uber's deals with Waymo gives them access to transport supply and technology they don't have on their own. And Waymo gets demand generation from Uber's platform to improve its expensive asset utilization. But over time, the economics of this partnership will depend on which capability is easier to replicate. On this front, I'd give the edge to Waymo, who are likely to be the only AV technology player for many years to come. It seems pretty clear that Waymo is testing the waters for now by splitting their operations between partnering with Uber in some cities and going their own way in others.
3. While picking winners in this space at this point is challenging, picking a loser is easy. The LAST thing rideshare drivers need at this point is more supply, and with AVs, the only question is how many, how fast!
The analysis is good. There is a reason Tesla is only at self driving level 2. While Waymo is at level 4. Even Mercedes Benz is at level 3. Tesla was supposed to pass Waymo in the self driving dust, even though Waymo had a 5 year head start. Tesla has over a million cars to Waymo's less than 1,000 cars on the road. That's a thousand times more data for its neural networks. But in Musk's infinite wisdom he removed radar and didn't use lidar. He said people don't use them so I don't need them. But people have a more developed neural network. He also wanted to cut costs, even though historically electronic device's costs have declined.
Intriguing analysis, but how is the situation affected by the fact that many rideshare drivers drive for both Uber and Lyft? And of course, most customers use both Uber and Lyft, and would add Waymo, when it becomes more widely available. Are there data on how many unique drivers drive for Uber and Lyft, and how many split their time between the two? In other words, is there double-counting in the estimates of the capacity of each platform?Likewise, how many riders are uniquely loyal to Uber or Lyft, and how many switch back and forth? And how big of a deal are the "frictions" that you mention (and which seem to annoy you with Uber and Lyft) to most customers -- Has anyone studied this issue? If these frictions are indeed significant to a large proportion of riders, then Waymo would offer a significant point of differentiation.
I am not sure the results are affected by how much multihoming is done by drivers. However, if you want to see how many do now, I have co-authored a paper exactly on this topic: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4411974 (TLDR, 40% never multi home. The rest do). How many customers multihome, you can read here. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4591826
Once Waymo’s AV’s are on Uber, and assuming price benefits continue to flow to customers, Uber will have a steeper internal competition between drivers and AVs. With driver earnings already a fraction of total ride cost, it would be interesting to see how the platform continues to stay lucrative for drivers.
There is a research paper on this topic as well. At least initially, they will have to pay the remaining drivers more to keep them on the platform (since drivers will know that within a few years they will be displaced).
The evolution of urban mobility will indeed be a fascinating business case playing out in plain sight over the next decade(s) to come. In addition to the dynamics you mention, I would add a few other considerations.
1. I don't see Tesla being a player for a meaningful future (say 10 years). for both technological and regulatory reasons. Let's not forget that Tesla's FSD is mired at Level 2 AV operation (and still has a high accident/fatality rate), while Waymo is operating safely at scale at Level 4 (and about to add highway operation capability). Yes, I know Mr. Musk has previously pulled innumerable rabbits out of his hat before. Still, the technological challenge of operating Level 4/5 AVs in dense cities like NY, Sao Paolo and Mumbai is a LOT harder than catching a reusable rocket on a launch gantry. The former has to cope with the extreme volatility of human behavior and unpredictably variable micro-level edge-case circumstances while the latter is "just" physics in a controlled operating environment.
On the regulatory front, we haven't even begun to sort through the insurance, risk management and operational implications of allowing individual vehicle owners the freedom and ability to determine urban mobility supply on a totally decentralized basis. Fortunately, the regulatory concerns are moot for the foreseeable future, given Tesla's technological challenges.
2. For now, both Uber and Waymo have unique competitive advantages that are valuable to each other. Uber's deals with Waymo gives them access to transport supply and technology they don't have on their own. And Waymo gets demand generation from Uber's platform to improve its expensive asset utilization. But over time, the economics of this partnership will depend on which capability is easier to replicate. On this front, I'd give the edge to Waymo, who are likely to be the only AV technology player for many years to come. It seems pretty clear that Waymo is testing the waters for now by splitting their operations between partnering with Uber in some cities and going their own way in others.
3. While picking winners in this space at this point is challenging, picking a loser is easy. The LAST thing rideshare drivers need at this point is more supply, and with AVs, the only question is how many, how fast!
Completely agree with everything!