Of course. Yes, of course. So I mean, we can give you our views, but I will tell you that probably the dominant thing to take away on that is a little too early to tell. And the reason I say that is because if you look at the markets where AVs are operational right now, you see a San Francisco, you see a Phoenix, you see in L.A., you see in Austin as an example. And the dynamics in each of these is slightly different. In some markets, AVs are actually priced as a premium product. In some markets, they're priced at a little bit of a discount. In most markets, the service levels are actually not as good as traditional rideshare, which you can understand why that might be that the ODDs are somewhat limited. And even when there's dense supply, those companies don't necessarily -- or it's really that company doesn't necessarily have the same sophistication with supply-demand balancing and so forth. So it's very hard to know over the long term, what's going to happen. Obviously, the bet that generally people are making is that over time this will increase supply overall and decrease cost to operating a network like this. And that seems like a reasonable assumption for obvious reasons, but that could be quite a long ways away. It really could. For example, insurance, people think a lot about insurance, got to [ph] be safe. That's true. But on the other hand, they're very expensive to repair. That's also true. They still do get an accident because other people cause those accidents sometimes. The utilization rates are very unknown. The profitability of these on a unit basis per car basis will have a lot to do with whether they're utilized 14 hours a day or 18 hours a day. So -- and then -- sorry, back to insurance, [indiscernible] I mentioned insurance companies tend to be quite conservative when it comes to pricing insurance, particularly early on, they say they like to have 5 to 7 years of data, which means -- and sort of I'd say their incentive is maybe even to overprice at the beginning, so they don't get caught upside down. So it might not be a very satisfying answer, but I think that's kind of the truth. Long term, I would expect again that it will be very sort of expansionary and maybe bring prices down. But I don't bet on that in the super short term. I think more interesting in the short term is just how differentiated they are from the existing offering and therefore, market expanding. On the issue of individual ownership and the particular Waymo Toyota thing, I won't probably mention or talk about too much. But I think what it does is it validates that there are going to be a couple of phases of AV adoption. The first, of course, is very pilot and trial, and that's obviously where we are now. Over time, there will be owners of fleets, right? And today, those owners typically are the people who are designing and building the tech, Waymo being obviously exhibit A, they design the tech, they buy the cars from Jaguar and then they own themselves. But over time, you wouldn't expect that necessarily. You would expect new companies like Marubeni and others and even companies like Lyft at a small scale to actually want to own the assets and manage the assets and so forth. But then I think in the medium and long term, you get to a very different and next level where individuals or maybe very small fleet owners of small groups of cars, own cars. And we're already planning for that very, very directly. So this is why I think our Mobileye partnership is so important. Again, Mobileye, they're the driver assistance technology that's in 800 different car models right now. So they have very good relationships with OEMs, but it's Level 2 kind of lane assist and so forth. But they have a very clear strategy and a very clear roadmap. And again, we're dependent on that roadmap as we roll out in Texas with them to get to Level 4 and Level 5. The idea there is that, that technology won't just be in bespoke cars, 1 or 2 OEMs or even 1 or 2 platforms within 1 or 2 OEMs, but rather ultimately universally available. And as I've said before, I think there will come a time, call it, the next 10 years, probably not much sooner than that, but 10 years, where buying a car without AV will be like buying a car without a radio. It be -- actually, I think a better analogy is buying a stick shift car. I think that's a better analogy. You can buy stick shift today. I happen to have one. They're awesome. And they're super fun to drive, but they're definitely the thing that you drive because you like to, not because you -- it may not be your best commuter option every single day. So I think it will be kind of like that. There will be manual cars just as there are today, but increasingly, AV will be everywhere. And in that world, this idea of lift Lyft ready, every car being Lyft ready so that when you buy it, you can put it on the Lyft platform and have it generate revenue, have it sort of self-monetized, you might say, becomes very appealing. So that's the world we're already building to. We know it's a long way away, but we also -- we have a very long-term view on AV. I think you have to. Again, for all the sort of short-term buzz there is and the flurry press releases and so forth, that's one thing. But I think sort of you have to have your eyes on the long term on this.