3 ms·

What do you mean? Cars are a marvelous solution to traveling a medium distance rapidly. They are ubiquitous as a result. I doubt we will ever have effective pu

What do you mean? Cars are a marvelous solution to traveling a medium distance rapidly. They are ubiquitous as a result. I doubt we will ever have effective public transit across the US. It’s just too big and most places have too low a population density to make it cost effective. And the price of convenience would be high. I used to be bullish about autonomous driving but no longer am. Ten years ago it felt like a much safer-than-human computer driver was within close reach. It no longer looks so rosy. I think Waymo’s approach will prove the most successful, and will end up with many of the same constraints as public transit. These companies have been going at it for a long time and have mostly demonstrated just how hard the problem really is. There is a decade of sunken cost for anyone leaving the market now. Would replacing human drivers be cost effective in any reality? I’m unsure.

danny_codes 3y ago
> It’s just too big and most places have too low a population density to make it cost effective. I hear this all the time and it still makes no sense at all. Population goes where transit allows population to go. Cars resulted in low population density. By getting rid of cars, you get rid of low population density. As for cost, there is no world in which every person owning an automobile is cheaper than public transit. It's not remotely close. Average cost of private car ownership in the US exceeds 10k a year. That's about 3.4T of private spending, or 14% of GDP. This does not include the cost of roads. A single year of private automobile spending would build 20k miles of high speed rail, even assuming costs were CA HSR high. I feel like the "low pop density & too big" argument is one of those things that only makes sense if you never consider second-order effects.
mikepurvis 3y ago
I'm hopeful that robotaxi services end up teaming up with mass transit agencies so you plug into your app "I'm going to X" and the options are: - point to point in a private vehicle: x time, $y - to the RT station: x * 1.2, y * 0.3 - to the bikeshare cluster: x * 1.6, y * 0.1 Governments have a role to play in creating the proper incentives here too, ensuring that congestion is properly paid for— every time any road is moving less than 80% of the speed limit that should be considered a failure of policy that it was too cheap to drive on it that day.
- vidarh 3y ago
  A pet idea of mine is app-driven robobus that scales in vehicle size all the way down to that robotaxi and where picking journey by app is encouraged because when you do there is a "wait time guarantee" for the bus service where holes are plugged with robotaxis/minbuses. Being able to more or less promise a maximum wait time and being able to dynamically scale a bus service in smaller seat increments without the cost per seat going through the roof (I looked up UK coach hire costs breakdowns, and when hiring a 70 seat coach the cost of the driver still added up to about 40% of the title cost including amortisation of the vehicle itself; drop vehicle size and presumably that goes much higher and the per seat costs accordingly), combined with being able to flexibly feed passengers to transit hubs in smaller seat increments on routes where demand can't fill a full size bus all the time could make transit services vastly more convenient.
tayo42 3y ago
We should frame public transportation as a defense issue and use that obscene budget to fund and subsidize it
- xnx 3y ago
  I don't think I follow. Rail lines are fixed targets and extremely vulnerable to attack. Cars can route around damaged roads and even drive off-road in some cases.
  - danny_codes 3y ago
    He's not saying it's actually for defense. He's saying we _say_ it is, and then magically there will be funding for it.
  - tayo42 3y ago
    The us spends a ton on defense, the money is there. im not a defense expert but im sure we can squint enough and figure out how to make good use of that money instead of spending it on wars
wombatpm 3y ago
I'm still bullish on automated long-haul trucking. I'd much rather share a road with an self-driving semi on I80 instead of a human driver hopped up on speed driving 14 hours a day.

danny_codes 3y ago

> It’s just too big and most places have too low a population density to make it cost effective. I hear this all the time and it still makes no sense at all. Population goes where transit allows population to go. Cars resulted in low population density. By getting rid of cars, you get rid of low population density. As for cost, there is no world in which every person owning an automobile is cheaper than public transit. It's not remotely close. Average cost of private car ownership in the US exceeds 10k a year. That's about 3.4T of private spending, or 14% of GDP. This does not include the cost of roads. A single year of private automobile spending would build 20k miles of high speed rail, even assuming costs were CA HSR high. I feel like the "low pop density & too big" argument is one of those things that only makes sense if you never consider second-order effects.

mikepurvis 3y ago

I'm hopeful that robotaxi services end up teaming up with mass transit agencies so you plug into your app "I'm going to X" and the options are: - point to point in a private vehicle: x time, $y - to the RT station: x * 1.2, y * 0.3 - to the bikeshare cluster: x * 1.6, y * 0.1 Governments have a role to play in creating the proper incentives here too, ensuring that congestion is properly paid for— every time any road is moving less than 80% of the speed limit that should be considered a failure of policy that it was too cheap to drive on it that day.

vidarh 3y ago
A pet idea of mine is app-driven robobus that scales in vehicle size all the way down to that robotaxi and where picking journey by app is encouraged because when you do there is a "wait time guarantee" for the bus service where holes are plugged with robotaxis/minbuses. Being able to more or less promise a maximum wait time and being able to dynamically scale a bus service in smaller seat increments without the cost per seat going through the roof (I looked up UK coach hire costs breakdowns, and when hiring a 70 seat coach the cost of the driver still added up to about 40% of the title cost including amortisation of the vehicle itself; drop vehicle size and presumably that goes much higher and the per seat costs accordingly), combined with being able to flexibly feed passengers to transit hubs in smaller seat increments on routes where demand can't fill a full size bus all the time could make transit services vastly more convenient.

tayo42 3y ago

We should frame public transportation as a defense issue and use that obscene budget to fund and subsidize it

xnx 3y ago
I don't think I follow. Rail lines are fixed targets and extremely vulnerable to attack. Cars can route around damaged roads and even drive off-road in some cases.
- danny_codes 3y ago
  He's not saying it's actually for defense. He's saying we _say_ it is, and then magically there will be funding for it.
- tayo42 3y ago
  The us spends a ton on defense, the money is there. im not a defense expert but im sure we can squint enough and figure out how to make good use of that money instead of spending it on wars

wombatpm 3y ago

I'm still bullish on automated long-haul trucking. I'd much rather share a road with an self-driving semi on I80 instead of a human driver hopped up on speed driving 14 hours a day.