Hacker News Reader

Comment by dalbasal | Hacker News Reader

One feature of the transport market (from a mile high view) is that demand is almost infinite.

If you can get 40kms in 45 minutes, then you can live within 40 miles of work and commute in that time. If you could go 400kms, then that becomes your commuting radius. We've seen people spread this way as earlier mechanized transport was invented. IE, average commute times didn't go down because of faster transport, distance traveled (KMs consumed) went up. Total commuting time can even increase as transport gets faster/cheaper.

If I could get to hong kong in 15 minutes, I might go there this afternoon for some dumplings.

That creates some strangeness in transport. You can't really solve transport the way people want, which is to make getting where they're going better/faster. If you could get there faster, more people would be going there (congestion) or maybe you would be living farther away.

This definitely doesn't mean it's not worth innovating, just that predicting how this will play out can be unintuitive.

View on HN

Yes, but it creates huge economic positive impacts

1. You can live in a less expensive area, (i.e. miles away), and work in a major center. This makes both possible of talent being able to live cheaper, and be able to work where they want (less geographical restrictions), which unlocks greater efficiency of the work place.

2. The alternative? Most people living with few tens of miles of workplaces, prices of living raise up of these hubs, you end up with both more expensive workforce, and less flexibility, and some people just decide to leave. The only winners in this scenario are current landlords.

Right now SF is experiencing scenario #2.

Basically once you have super fast transportation, it relieves pressure to housing prices on the work hubs/city centers, and the values of homes where the high speed train/hyperloop/whatever stops increase to match the new accessibility.

Real Scenarios: You can comfortably live in Baltimore or Philadelphia (2hrs of driving currently) and be able to work in NYC if this commute was cut to 40mins or less.

It is a win win scenario, as all these cities become part of one single hub.

That's a very rosy view, only looking at the positive side of the coin and ignoring the negative side.

>1. You can live in a less expensive area, (i.e. miles away), and work in a major center.

...causing property values to rise in the "less expensive area" to match those in the urban center minus the cost of transport. Ultimately you're gentrifying the current residents out of their homes, causing more urban sprawl, and increasing land use pressure on agricultural land and wilderness.

>Basically once you have super fast transportation, it relieves pressure to housing prices on the work hubs/city centers

...shifting that pressure to surrounding rural areas. TANSTAAFL.

You talk about "huge economic positive impacts," but it sounds awfully like wealth transfer from rural people (which we're apparently ignoring?) to urban people. Given the audience on HN this bias is unsurprising, but still disappointingly common.

> ...causing property values to rise in the "less expensive area" to match those in the urban center minus the cost of transport.

This is a valid concern made invalid by the speed of the Hyperloop. If we estimate very conservatively that the Hyperloop can cut the average commute time 4x, then the radius that a city worker can live increases 4x. However, while the radius increases 4x, the area increases 16x (A=pi*r^2: area increases exponentially as a function of r). Thus, the rise in property values in surrounding areas would be distributed over a large area so as not to make life unsustainable for those already there.

Quickly, that's not exponential, it's quadratic.

More importantly though, it doesn't fix the travel time issue that well. It only speeds up a particular section of the journey and only between two specific points. You would need vast amounts spreading out with lots of stops to get close to a squaring.

The housing pressure from San Francisco looks like it currently extends at least as far as Gilroy to me (houses going for 1/2 million dollars).

That's without it being a viable commute to the peninsula.

http://www.caltrain.com/schedules/weekdaytimetable.html

It's presumably a viable commute to the south bay?

Why commute ? Remote working is a thing and everything the topic comes up, something that a significant part of HN is already enjoying. No need to pour billions in public transport.

Also companies have proven that they are more than willing to outsource everywhere, hiring people anywhere. There are reason they still hire in very expensive place like SF. You bet that it is a problem that will go away at some point, and very likely a significant part of HN readership is working hard solving it.

You need to realise that moving commuter to the city centre is a race to the bottom. Stuff like hyperloop are a perfectly fine way to spend private investment money, but as far as public money is concerned, that would be better to invest it in area that look beyond a world where physical location is an employee primary competitive advantage.

Well... observable trends atm is that physical location still matters. Remote work is more common, but at the same time major centres for tech industries (SF famously, but also others around the world) are pulling in more residents who are going there for work

Getting people to workplaces is a thing we need in 2017, and I expect in 2027. Beyond that, who knows.

> observable trends atm is that physical location still matters

Sure, we're still in the shallow part of the disruption curve. The real world (established player) has two features that Skype/Hangouts/Slack/VR (disruptive player) can't do yet:

- eye contact

- a semi-conscious background channel

Group video calls are confusing without eye contact. You can't really pay half attention to Slack. You read, or you don't read.

But within 5 years or so you should be able to buy a $400 VR headset which lets you make eye contact with someone, and place a semi-interesting conversation between coworkers in the background behind your virtual desk.

And of course the VR workplace has many benefits which will never be replicated by a real world workplace. It's a classic disruptive technology. Small market now. Will eat almost all of office space within 20 years.

I would be lonely doing remote work. I prefer to go to an office and interact with people every day. Sometimes I don't want to interact with them, but I wouldn't want to spend 8 hours a day alone in my house.

Coworking spaces might work. I haven't tried it, but I imagine its different interacting with people who aren't relevant to what you're doing all day.

That sounds suspiciously close to "I go to work to drown my loneliness". Might have not been your intention but it definitely sounded that way.

To me, work is work -- no attachments. I happen to like several of my remote colleagues quite a bit but I am doing just fine in my life if I don't chat with them for a month or two.

You won't be alone, you'll be in a virtual office overlooking the Swiss alps, and all of your coworkers will have a mute button.

> Also companies have proven that they are more than willing to outsource everywhere, hiring people anywhere.

Debatable. Googlers complain of pressure to relocate to Mountain View (despite using video even for intra-campus meetings), Mayer shut down telecommuting at Yahoo, Facebook built the biggest open office in the world.

With great collaborative tools, a greater concern for carbon emissions, and the cost-saving benefits of working remotely, daily commutes to the office will decline.

..in theory. In practice, SF, can probably grow it's commuting catchment area (supply) by quite a lot before housing prices start to give. I would predict that you'd just end up with a bigger SF.

#1 most of the time isn't true because public transportation is very poor in those areas.

One feature of classical mechanics is that the world must abide by it. With transportation, the vehicle must start and stop at a speed of zero. The energy required to accelerate the vehicle to its top speed and then decelerate it to back to zero increases quadratically as the top speed increases.

Another feature of classical mechanics is, of course, that Aristotelian concept of kenophobia, or horror vacui, or resintenza del vacuo, or natura abhorret vaccum, or nature abhors a vacuum. Of course we now know we can produce vacuums these days... but it's not easy to make one. The largest is in Sandusky, OH, and it's 122 feet high and 100 feet wide. It takes 8-12 hours to achieve vacuum conditions. https://en.wikipedia.org/wiki/Space_Power_Facility

> The energy required to accelerate the vehicle to its top speed and then decelerate it to back to zero increases quadratically as the top speed increases.

What? No. You can, in fact, recover a substantial portion of the energy used to accelerate the vehicle when you decelerate it (quite a lot of it, if it's running in a vacuum with minimal frictional losses). The track would act as a linear motor during acceleration, and a linear generator during deceleration.

> Of course we now know we can produce vacuums these days... but it's not easy to make one.

You don't need a vacuum anywhere near as hard as the Sandusky facility for a transportation system. That operates at 10e-6 torr, while the Hyperloop will operate at 100 Pascals (about 0.75 torr, or 75,000 times higher).

That's a supply problem.

The energy required only grows quadratically if you stay on one technology.

Simultaneously switching from internal combustion to linear electric motors with efficient regen. braking and bumping the speed up 10x turns out to be a wash.

To clarify, you are stipulating that electric engines and regenerative breaking are ~100 times more efficient than a combustion engine? Because mv^2 means a 10 fold increase in speed is a 100 fold increase in kinetic energy.

Thanks for the catch, I wrote that comment a little too flippantly and left out important factors, the motors/regen alone certainly don't get you there.

Adding in the reduced drag of levitation vs soft tires combined with central control to avoid stop-n-go intersections/traffic gets you to 100x more efficient.

Edit: Also, I should just drop this link to some envelope math I did last Fall:

https://medium.com/@nickparkerprint/why-im-still-excited-abo...

And since I didn't think of this either before the first comment: I'm currently an employee of Hyperloop One, but none of the statements made in this thread are at all based on H1's design or proprietary info. They're based on the envelope math above, which was done before I started working there. I don't actually know H1's efficiency numbers because it's not my department.

> With transportation, the vehicle must start and stop at a speed of zero. The energy required to accelerate the vehicle to its top speed and then decelerate it to back to zero increases quadratically as the top speed increases.

Don't be so sure. The vehicle could move continuously at cruising speed while individual people are brought to/from it and then "dock".

Those people still have to accelerate / decelerate.

> If I could get to hong kong in 15 minutes, I might go there this afternoon for some dumplings.

And that's a good thing, because it increases the amount of demand available to businesses. The more potential customers you have, the more product you can make to sell to that market, and the more product you can make, the better your economies of scale.

The large population of New York is what allows it to offer so many niche and specialty stores, which would otherwise struggle to stay in business.

counterpoint: it's also a bad thing because society functioning now becomes more reliant on this transportation method.

It destroys "walkable" spaces and encourages centralization of services, when a more uniform spread of services across space is in fact healthier.

I heard of a study recently that showed that cities in Japan that ended up getting Shinkansen stations ended up almost always shrinking, because people got used to going to Tokyo all the time and ended up just moving there. You see similar centralization in places like France (everything goes through Paris)

I realize your point but food deserts exist for a reason.

> a study recently that showed that cities in Japan that ended up getting Shinkansen stations ended up almost always shrinking

I welcome you linking to this study, because a naive explanation for parts of Japan shrinking would be Japan's overall population decline (i.e. shrinking demographics). So unless the study controlled for that, I'm rather disinclined to think that the Shinkansen by itself caused areas outside Tokyo to shrink.

If anything, centralization happens because of the difficulty of transportation. Highways were what drove "white flight" to the American suburbs. Jammed highways and hellish commutes are driving gentrification.

And centralization improves walkability, if anything. If I need to take a train to get into the city, then I need the city to be walkable after I get there, because I don't have a car with me anymore.

You see the same in London too, where even within London, boroughs that'd be viable large cities of their own if they were further separated from London, has big gaps in various types of things because you can just go in to the centre.

E.g. nightlife is a typical one, where there are a few centres in London (and they change over time), and clubs and bars outside of those regions suffer compared to even smaller towns elsewhere.

On one hand it's great - you can head where the action is. On the other hand it means a night out might easily be an hour travel each way and/or expensive tax rides, instead of a 20 minute walk.

This is so counter intuitive. This is the exact opposite of suburbia. Faster and cheaper transport led to the growth of suburbs. Can the movement of people to Tokyo be attributed just to the faster trains ? Can you please share a link to the study if you have it handy.

Shinkansen isn't really priced to be a commuter rail service. Nagoya/Tokyo is roughly $90 each way.

Employers subsidize it.

Japanese salaryman employment is like civil service employment. Mediocre pay + crazy benefits.

They tend not to pay fully for shinkansen tickets (monthly commuter tickets end up being a lot of money). Though some cities tend to subsidize them.

OTOH, it reduces the amount of such businesses, since each one now has to compete with many more for the same customer base. In the limit (cheap teleporters) this leads to what we have on the Internet - winner-takes-all global competition.

On the contrary, such competition drives business to hyper-specialize, in order to carve out its own market niche. There was so little competition in Ye Olde Medieval European Village that you didn't have a choice what kind of meat or metal you could buy - there was a guild-protected artisan, and what he made was what was available for sale. Now you have a choice between organic and free-range (usually both) and between Game of Thrones replicas and Lord of the Rings replicas.

Yes, but the point was that if you could get to Hong Kong in 15 minutes for dumplings, you couldn't get to Hong Kong in 15 minutes because everyone else was going there, too.

And some of us don't live in New York because of the friction involved.

The point (at least my point) is that if you could go to Hong Kong in 15m, we would be consuming a lot of "transport," thousands of times more KMs than we currently travel for dumplings or work or whatever. Whatever teleportation technology is responsible, needs to be almost unlimited in its capacity because our demand is almost unlimited.

When supply is limited (teleportation wormhole congestion) and demand is unlimited, something will negotiate the difference. This can be prices or congestion.

Are you arguing that nobody goes there anymore because it is too crowded?

If you aren't too pedantic about the definition of nobody it actually makes sense though. In this case "nobody" could be 5 billion+ people not going because it's too crowded, yet there would still be large crowd.

You're rubric for that which is good needs to be expanded beyond just "more money for businesses".

I live in London, and this is a pet peeve of mine. Central London traffic moves slowly above ground and the trains are usually full. So of course we need more trains? Billions gets poured into it.

But like you I don't think it will help.

Instead I wish government here would stop all expansion of Central London lines. Very publicly. And instead focus on picking, say, a dozen towns near the motorways and rail lines surrounding London, and expanding the connections between them and further out. Especially towns where there'd be little resistance to high density development.

Instead we get high speed links to other cities so people can travel from even further out to the centre of London in less time.

E.g. I live in a suburb where better train connections along the edge of London + further out could easily make it a viable commuter destination for 4-5m people where it's now viable for about 2m. In some cases the rail is already there, but there just aren't any direct trains. There are many similar opportunities. Developing satellites like this would be so much better than the relentless focus on the centre.

And there are many hubs like London around the world where we're just shooting ourselves in the foot by seeing congestion and thinking the problem is lack of roads, rather than seeing congestion and thinking it reflects a lack of alternative destinations.

demand is almost infinite

Almost infinite but still very much finite. If you could get anywhere in the world in seconds, Star Trek transporter style, then demand would level off. The number of people in the world is finite. Once everyone can get where they need to be as easy as sending a text message, we will have different problems to solve.

If you could get anywhere in the world in seconds, Star Trek transporter style, then demand would level off.

Hell no. People raised on "jump cut" vlog culture will just start jump cutting in real life. (Before anyone takes offense, they should first read Larry Niven's seminal articles about how Sci-fi authors should extrapolate sociological issues, using teleportation as his prime example.)

I'm interested in these articles you mention, but am having trouble finding out more. Do you remember what they were called, or any details about their publication?

I remember one particular sci-fi story (I don't remember right now if it was by him though). In it they have booth to booth teleporters that can take you virtually anywhere on the planet that's been hooked up.

One of the consequences of the technology is that immediately spectator value goes off the chart. As soon as people hear about something interesting happening anywhere then there is an an influx of a few million people playing spectator, which as you can imagine, has some consequences.

They were published in anthologies of his short stories and articles. EDIT: I believe it's in the anthology All the Myriad Ways.

> If I could get to hong kong in 15 minutes, I might go there this afternoon for some dumplings.

I doubt you would because the cost of the trip would make that the most expensive food you ever ate.

A round trip economy ticket from New York to Hong Kong [1] is about $700. Kayak shows decent hotels at about $200 a night. Let's say 2 nights. That's $1,100 for travel and board.

To match the $3,000 a romantic dinner for two costs at Thomas Keller's Per Se, I'd have to spend almost $1,900 on dumplings and knick knacks.

[1] On Delta, about a month out

You're looking at closer to $1500 total for dinner for two with wine at Per Se, unless you're ordering something extra-fancy.

That's true.

But most things people consume don't scale that way. You only need so much clothes or square footage of house. Some people might get extreme given infinite amounts of stuff, but most won't have that much more of it than many people have currently.

Transpo wonks call this "induced demand" - the component of transportation demand caused by people reacting to the availability of transportation. Every infrastructure improvement induces a certain amount of demand, including incremental ones like adding lanes to roads. It's the reason new lane additions always seem to reach a point of being congested, faster than anticipated.

In this way transportation spending is both the purported cure and the cause of congestion, and fits the definition of an addiction. See my comment from a few days ago: https://news.ycombinator.com/item?id=14850885

It's 100% observable. I used to work for a regional transportation council that made high level decisions for where to alleviate congestion. There were many hundred-million dollar projects that counterintuitively increased congestion, because the increased availability created by the infrastructure caused even more people to move to the area. After the original project was completed, decreased congestion would be observed, but 2-3 years later it would be even more congested than before the project began.

One feature of the transport market (from a mile high view) is that demand is almost infinite.

It seems this is also true for cars and trucks. No matter how many roads we build, there is congestion. I think automated cars are going to show that one of the limiting factors to how much our society uses such transportation is the supply of piloting labor.

Also to chime in: It's not literally infinite. It's just much higher than the current economy can supply.

This is known as the Pigou-Knight-Downs Paradox (or Downs-Thomson Paradox). Building new roads creates its own demand. Congestion is a function of the speed of alternative modes of transport, rather than road capacity. This is because when you build a new road that improves journey time, it becomes more appealing to drive. Congestion increases to the point when it becomes quicker to get public transport instead of driving. If you want to cut congestion, don't build roads, build railways.

You could also just legislate that every time one uses a car you have to stand in a shaking room packed with people for half and hour. I bet that would really cut down on congestion. If not, just up the wait time in the hot box until it does. People, could of course, would have the option pay for someone to do their waiting for them.

And self-driving cars may increase congestion yet.

Whatever speed gains self-driving cars may attain by reducing time spent at a (now deprecated) traffic light or at an intersection/yield sign, people may just make more trips because road conditions have temporarily improved, thus cancelling that benefit.

Say you decide not to drive to work but instead take the rail because you don't want to deal with the hassle of traffic. Maybe the rail has its downsides too, but they are acceptable to you. Now if self-driving cars makes highway congestion to work better, you may just be that person who switches from rail to car.

I think for it (demand) is very high, but it's not really possible to say for sure without supplying a lot more, free piloting can give us. I think infrastructure will continue to be the limiting factor.

* I think infrastructure will continue to be the limiting factor.*

Autopilots with superhuman reflexes and "telepathic" sharing of sensor data/intelligence could be used to increase the carrying capacity of roads by almost an order of magnitude. Photos of roads indicate that we are only using 1/10th of them to carry cars, spatially speaking.

> No matter how many roads we build, there is congestion.

Where? When?

How is demand infinte? Demand is a function of price. We could fly to London from NYC on supersonic jets for decades, but we decided that it's too expensive. How is transportation different from any other service industry?

Right, but that can't be done by fixed infrastructure like rails. What happens to area as radius goes up? What happens to population density and demand?

very interesting point and analysis!

Reply on news.ycombinator.com