Hacker News Reader

Comment by comte7092 | Hacker News Reader

At the end of the day that isn’t a technical problem but a unit economics one.

Removing the driver from a taxi doesn’t bring down costs that much. Self driving cars aren’t going to change the uber/taxi model at a fundamental level.

They have a finite fleet that they need to deploy. Urban centers mean that fleet utilization is high, and relatively less time and miles are spent driving with no one on board. In rural areas with little demand they will sit empty or have to drive empty for many more miles to their pickups. It just isn’t profitable to use your fleet that way no matter what you do.

View on HN

Hmm, I'm not sure about that - fully autonomous taxis wouldn't be subject to the limitations of human drivers in terms of availability / reliability / endurance. You could ostensibly leave a few taxis around to service otherwise underserved areas and have them run without having to secure a driver each time.

That being said, there is still the cost of maintenance and cleanup, but that can be mitigated (the taxis for five towns could drive to one centralised depot, maintenance can be scheduled to maximise operational time, and eventually all of this can be automated, too)

I don't know if that's how things will work out just yet, but it seems like a possible future based on Waymo's current operational strategy.

> You could ostensibly leave a few taxis around to service otherwise underserved areas and have them run without having to secure a driver each time.

I think you’re dramatically overestimating how much of a barrier obtaining a driver is here. The primary cost is opportunity cost of the capital that isn’t being utilized. Not having to have a driver doesn’t somehow make it so you can infinitely provision a fleet.

They won’t have to pay for the driver when idle, but owning cars ties up capital and the fewer rides they do, the longer it takes to pay off. This isn’t specific to cars - all capital equipment works that way. Lower utilization is sometimes unavoidable, but it still means less revenue which can be the difference between a profit and a loss.

How much this matters depends on the price of the car. We don’t know how much a Waymo costs, but they’re probably not cheap.

To be profitable with lower utilization, they’ll need to work on reducing how much each car costs somehow.

> To be profitable with lower utilization, they’ll need to work on reducing how much each car costs somehow.

Definitely. Their custom vehicle had optimizations for cost, but seems to be on hold due to tariffs.

Waymo also has the option to drop prices lower than Uber/Lyft when vehicles are unutilized, though they still need to stay above their per-mile depreciation and operating costs.

> Waymo also has the option to drop prices lower than Uber/Lyft when vehicles are unutilized

I think that’s an unproven assumption.

There’s certainly reason to believe it to be true of course, but uber and Lyft are already capturing upwards of 50% of the fares for each ride, and that’s without the capital costs on their books. Removing the driver from the equation can’t lead to much more than that 50% (realistically much less) margin.

Going from charging $10 to $5 isn’t going to make rides suddenly materialize. Especially in rural areas there are just times that people aren’t going to be looking to go anywhere, and wait time becomes far more of a factor that raw costs.

> There’s certainly reason to believe it to be true of course, but uber and Lyft are already capturing upwards of 50% of the fares for each ride,

That's not true. If you check Uber's Q3 financials, gross bookings for "Mobility" were $21B while revenue was $6.5B. That's way lower than 50%.

> but uber and Lyft are already capturing upwards of 50% of the fares for each ride

I'm not sure if that's an accurate number, but I have seen a lot of complaints from drivers that they're getting a far lower share of the trip revenue than they used to. It's pretty remarkable that in a competitive market where Uber and Lyft are almost perfect substitutes for each other and charge almost exactly the same prices, that they're able to maintain these gross margins.

Theres a bunch of factors that will mess with your intuition here: 1) ride hail demand as significant spikes in usage during morning and evening rush hours AND it has a fairly strong seasonal trend depending on geo. 2) Insurance is also a big expense and for large operations like this is priced per mile or per operating hour, having more deadhead time means a higher loss to insurance. 3) People are very sensitive to wait times AND reliability. The desire to use the service drops a ton when wait times are greater than 10 minutes or if you're consistently not able to find a ride. Could waymo support less dense suburbs now? Maybe at certain off peak hours, but the economics and product experience are difficult.

> Removing the driver from a taxi doesn’t bring down costs that much

The average pay for a gig driver $18/hr. So for your typical 15-minute ride, that adds $4.50 to it.

Let's say that 15 minute ride is 10 miles. Average. Uber rates are about $1.50 a mile, so that ride is $15.

Therefore, the driver costs almost a third of the cost of the fare.

Waymo's operational cost per mile, however, should be much lower than a regular driver because they will pay lower bulk rates for energy (already much cheaper because it's mostly off peak electricity instead of gasoline) and maintenance (standardized vehicles with highly controlled driving patterns and pre-negotiated repair contracts).

> Removing the driver from a taxi doesn’t bring down costs that much

A driver's salary costs as much as a new car. Every year.

That doesn't follow from the facts. Rideshare drivers get around 50% of the fare (though this seems to vary from 25%-75%). And many riders tip 10-20% of the total. Most of that cost goes away if a single operator is monitoring ~10 cars. The tip goes away entirely.

An average per-trip reduction of ~50% changes the economics entirely.

The driver's 50% pays for depreciation, insurance, car maintenance, and cleaning, all of which will need to be paid for no matter what

I know that rideshare workers have a raw deal, but they are absolutely not paying all of their earnings towards cleaning their car and insurance give me a break. Fulltime drivers are making a living. A robot literally requires 100% less costs for living. Plus no tip.

Removing the driver does allow for single occupant models that could be significantly cheaper (reduced materials, smaller battery - assuming they'll all be EVS).

It will be interesting to see how things develop once the driver is no longer required and cost is the most important factor (after safety). Exciting times!

Taxis today are already far far larger than they need to be for two people.

Logistically you need the flexibility of having more seats available. If you’re in a rural area and need to transport a family are you going to send 4 vehicles separately?

Ironically it’s probably urban areas where single occupancy vehicles make the most sense, given that there’s always going to be sufficient demand to allow for more specialization in vehicles for different use cases.

> If you’re in a rural area and need to transport a family are you going to send 4 vehicles separately?

Probably still years off, but Waymo will probably have a library of vehicles ranging from 2 seats to 20.

We already have single occupant vehicles that are low cost and fun to use: bicycles and e-bikes. They are very popular in areas where safe infrastructure is available.

I'm already an avid cyclist throughout the year, but there's a real drop in the number of people cycling during the winter months. If we can get people in AVs I think this will be a real positive for cyclist for the following reasons:

1. Reduced curb space dedicated to parking. If you don't need to come back to the same vehicle you only need space to be picked up/dropped off, reducing the amount of parking spaces needed. This space could be used for separated bike lanes. 2. Safer - This is still an unknown but data looks good atm [1]. It would be even better if AVs could be design to prevent cyclists being doored that would be amazing.

[1] https://arstechnica.com/cars/2023/12/human-drivers-crash-a-l...

I envy your optimism.

Self-driving cars will eventually lower the total cost of driving and it will allow for longer commutes as people will be able to either sleep or do some work while in the vehicle.

The inevitable consequence of that is an increase of car traffic, which means more congestion, noise and air pollution (tires and brake pads). We can't know whether the theoretically lower collisions per distance traveled will translate into lower actual injuries until we know how much the distance traveled will increase.

Most importantly, the more people rely on a particular form of transport, the more they will vote to facilitate it, via more lanes, more highways, more forgiving legislation, etc.

I would rather see more active transportation and more efficient forms of transportation. Four-wheel single occupancy vehicles are just about the worst option of all.

I can’t believe the level of ignorance I’m seeing in these comments.

The driver is THE overwhelming cost of a taxi/uber. What are you talking about?

Your problem is you’re not seeing past the costs of a Waymo vehicle with all its sensors and LiDAR, plus all the costs of keeping high definition maps updated, and having teleoperaters on hand.

Tesla doesn’t have those costs. Their FSD version 13 already drives close to a Waymo of not better in some circumstances. its a done deal, Tesla has won this game

Reply on news.ycombinator.com