Hacker News Reader

Comment by spacKingChamath | Hacker News Reader

There is a strong monetary incentive to make sure every car has a high utilization time. The cars driving around empty are there so they can collect data to prove the safety. No one is targeting transit. Autonomous cars will also have a place in transit. Its highly likely one day you will be relying on an autonomous car to get you somewhere for cheaper than transit does today. AVs have real potential to get people to forgo personal car ownership which actually will reduce traffic, emissions and restore millions of sqft in parking spaces to be reclaimed for parks and housing.

View on HN

We make special lanes for cars with multiple passengers b/c we think driving alone contributes to congestion, emissions etc. Driving with 0 passengers might be best for the companies, but for the same reasons, is not in the public interest. I think it's entirely appropriate for the public to place stringent restrictions on rides which do not move people or goods usefully. If you need high utilization, figure out deliveries and freight. Do _something _ useful.

It might be cool if someone developed a human-like robot that walked and talked independently ... But not cool if that robot walks in crowded places, doing nothing in particular, and stopping inexplicably. I would expect it to be shoved down a stairwell on day one. It just turns out cars are a little more challenging to shove.

There are only a couple hundred AV cars on the road. They are not causing the level of congestion you are trying to suggest.

Remember that these companies cannot earn money from rides right now because the CPUC has not given them the permit to do so. Once that permit is granted, the monetary incentive will be very strong. There is cost to the company for also operating an empty vehicle that they will not want to pay either. I think the incentives are aligned on both sides.

I think it varies a lot by neighborhood though. In some areas you see them a lot. Perhaps there are specific routes or locations that are considered high value as test cases? The density is high enough that as a pedestrian and casual observer it's easy to see them do stupid stuff a human would know not to do with some regularity.

Sure it makes sense that density per neighborhood would vary for testing and in the future will vary by demand.

Any new technology is going to have challenges as it scales. I think the expectation that the system is only used in public once its perfect is extremely ideal. The systems are doing really well at being safe (there are no human deaths linked to AVs nor any severe accidents despite over 1 million + miles driven).

Do you think people living in the places that are disproportionately impacted, who are being used as a test course, and who don't receive benefit in exchange for inconveniences imposed have a reason to be upset?

I think a small, non-destructive direct action, social media push and telling people they can make a public comment on official proceedings is reasonable and appropriate.

> There is a strong monetary incentive to make sure every car has a high utilization time

Not based on real utilization of cars today (5% utilization, 95% parked)

There is a stronger monetary incentive to sell at least one car to every single person :)

> Restore millions of sqft in parking spaces to be reclaimed for parks and housing.

That was also my hope. I don't think it'll happen. People will use autonomous cars like they use uber/lyft/taxis but they will still have their personal cars... Those will need the same amount of parking, or more given electrification and long charge times.

Because if they don't have their own vehicle then prices will rise for using uber/lyft/taxis.

High utilization time is irrelevant; autonomous vehicles further encourage low-occupancy vehicle use (both additional trips and trips that are converted from walking/bicycling/mass transit), and should not have 'a place in transit' except in limited cases like, for example, rural transportation for senior citizens. They have no place in urban environs; there's no need for them beyond profit.

Ride-share vehicles crippled the world's cities by causing a massive increase in low-occupancy vehicle trips, increasing vehicle ownership, etc.

> No one is targeting transit.

Uh, there's been a concerted effort since the 1940's by the automotive industry to kill public transit, but ok.

Whether they intend to or not is irrelevant; these self-driving-car companies are clogging up our roads because apparently it's acceptable to "fail fast"

https://insideevs.com/news/670739/cruise-robotaxi-stuck-musk...

https://insideevs.com/news/625602/gm-cruise-chevy-bolt-gets-...

https://www.theverge.com/2021/10/14/22726534/waymo-autonomou...

https://www.theverge.com/2021/5/14/22436584/waymo-driverless...

You won't be satisfied until you accept that the public has demonstrated that they want low occupancy vehicles by using there dollars.

People enjoy the privacy, flexibility, convenience, and safety of having a vehicle space to themselves vs public transport. Even in economies which have heavily incentivized public transport (Singapore, Japan, Hong Kong), people still pay large costs to get access to the incentives I mentioned above. AVs are much better than personal car ownership and will play a big role in improving transportation.

I get throwing conjecture out there makes it seem like you found a real problem, but you have no solution. Even if you increased public transit with more rail lines and busses, you can't solve the privacy, convenience and safety factor with something that is public. There a reason we have locks on our doors and don't share our homes, even though most of our home is empty and we can only occupy one room at a time.

We need real solutions (which AVs are a part of) and not people swooning over idealisms.

> AVs are much better than personal car ownership and will play a big role in improving transportation.

But again what makes you think that AVs won't be personal cars with the same ownership model? GM is suddenly going to be cool with selling 1/10000th of the cars they sell today?

There is more money longer term in a complete AV system then in personal car ownership. You can see it in GM market capitalization today. They are worth ~30 billion while Cruise is estimated to also be worth 30 billion. Making money off each mile is way more lucrative then selling a vehicles most people don't update for 3-7 years at a time. And it will be better for the consumer because it means more miles per vehicles which = less total vehicles.

> You won't be satisfied until you accept that the public has demonstrated that they want low occupancy vehicles by using there dollar

What the public wants is irrelevant; the public wants to eat its cake and have it as well, all the while riding a magical flying unicorn. Low occupancy vehicle use in urban areas is not sustainable, and never has been. There simply isn't enough road space. In the space of 2-3 cars which likely only contain 2-3 people, a bus can carry fifty or more people: https://danielbowen.com/2012/09/19/road-space-photo/

...and you can't build your way out of the problem, because capacity does not increase anywhere near a linear rate as lanes are added: https://usa.streetsblog.org/2017/06/21/the-science-is-clear-...

The problem is so bad that SimCity creators had to massively fudge road capacity and parking, pretending that there are massive underground parking lots everywhere: https://www.reddit.com/r/fuckcars/comments/x6ao05/simcitys_c...

Climate emissions goals aren't achievable without a significant curtailing in low occupancy vehicle use, as well. The most efficient production EV, a Model 3, gets about 4 miles per kWHr. An e-bike...even a cargo e-bike...will get around 60 miles per kWHr.

Further: we can't afford it. At least in the US, a huge amount of infrastructure hasn't received the maintenance funding it needs, and we vastly overbuilt our road network - paving to everyone's driveway and building more and more roads as traffic increased - without thinking whether the long-term expenses were sustainable. In the near future, bridges are going to start collapsing because nobody wanted to pay for the upkeep they needed and now we can't pay for it.

> We need real solutions (which AVs are a part of) and not people swooning over idealisms.

Aside from the fact that our current road network is based on not just idealism but outright sticking one's head in the sand and not paying the maintenance costs...

There are "real solutions" working just fine in most of Europe and especially the Netherlands and Denmark.

> AVs are much better than personal car ownership and will play a big role in improving transportation.

The only thing AVs "improve" is eliminating labor costs and a potential increase in safety. They address none of the congestion and energy efficiency problems of low occupancy vehicle use, and by reducing the cost, the technology either increases profits or lowers the cost, both of which will lead to their greater use.

Safety isn't relevant in this discussion because people fear crime and rubbing elbows with smelly people who don't look like them. The reality is that public transit is orders of magnitude safer than low occupancy vehicle travel because collisions are far more common than crime, and public transit drivers are involved in collisions with pedestrians and cyclists far less than other drivers.

Reply on news.ycombinator.com