Um...

...Listen, Amy, it's nuthin' personal...

But no. This critique pulls it apart with salad tongs.

Elon Musk's greatest gift is his political savvy: His talent for lining up government money for whatever he wants to do. And Jerry Brown has already done that for California trains. They're peas in a pod, and both essentially corrupt. There should be no new train lines built between Los Angeles and San Francisco, not in the decades when Google is perfecting driverless cars.

(Link via Cowen.)

And not when the actual need to be somewhere to do administrative jobs goes down over time.

"While government champions the "high speed" train up and down California that won't actually be high in anything but continuing cost overrun, actual transportation innovation comes out of people in private business."

Not this one - and I blame public education. Anyone should notice that trains have to stop to get people on and off them. Completely aside from issues with rights of way, this completely kills any speed claims made for any transportation system. The laws of physics rule here.

Maybe Gruber can push this...

I have always been fascinated with the idea that Americans have some inner need to be somewhere else. It might be the same sort of mentality that causes them to ruin one location, move to another to "get away from it" - then ruin that one by doing the exact same things they did at the previous location.

That's not intended to harsh anyone's enthusiasm for serious innovation. But the Levy critique has some seriously good thoughts about the nature of scientific inquiry in our times.

In math, one common sanity check on a result is, “does it prove too much?” If my ten-page paper proves a result that implies a famous open problem, then either my paper is wrong or I’ve proved the famous open problem, and it’s up to me to take extra care to make sure I did not miss anything. Most people in this situation do this extra step and then realize that they were subtly wrong. If a famous question could be solved in ten pages, it probably wouldn’t still be open. The same is even true in undergrad-level proof classes: if your homework answer proves things that are too strong, you’ve almost certainly made a mistake.

In the 1980's, a few months after I learned what a matrix in linear programming even was (roughly), a brilliant young mathematician came up with a new way to make it happen. "Aha!," I thought. "I'll try to figure out how it works!"

I never had a prayer. All the easy stuff, all the important ideas accessible by lesser minds (if only through metaphor) has been done, probably centuries ago.

You heard about this, right? From now on, innovation will not be a casual thing.

Public enthusiasm for disruptive excellence is precious gunpowder, and we need to keep it dry.

My favorite article about Elon Musk comes for his first ex-wife. It's not the article in and of itself that I loved. It was a comment from a reader.

"Who knew the marriage of self-absorbed pseudo novelist marrying an anal retentive penny pinching Austistic wouldn't work out?"

(That wasn't the comment verbatim but it's close enough).

Pipelines for people. We can get from Los Angeles to San Francisco nearly as fast, now. I ignore the security theater, and the difficulty of getting to the airport because those will not magically disappear with tubes. I don't see the economics working,

Easy innovation is a periodic thing Crid. Essentially, some lunatic comes up with a useful idea. This makes rapid innovation possible as people explore this new way of doing things. Eventually things settle down and innovation slows. And then some lunatic comes up with a new useful idea.

This is one of the reasons physicists and mathematicians who are brilliant are also often described as young. Once you hit 35 in those fields you probably aren't making any ground breaking discoveries. The material sciences is the exact opposite. Few people make groundbreaking discoveries there under 50.

And to be on topic, I don't see how a useless tube is better than a useless train.

It's a grand idea and I'd ride one for fun.
Maybe a better place to showcase the idea is along Highway I-15 from San Bernadino to Las Vegas. It's mostly passengers - not cargo, especially on weekends when the drive can be 18 hours round-trip because of traffic. And Vegas is a theme park now. A hyperloop into town would be part of the fun.
They could build it for less and sell tickets to L.A. residents who want to get there and back fast for their weekend off from work.

And to be on topic, I don't see how a useless tube is better than a useless train.

Well, it won't be nearly as expensive. So there is that. And there likely be fewer accounting irregularities, political kick backs, and union shake downs.

The elephant in the room is thus: how much of a subsidy does each of the modes of transportation require to be "profitable"? passenger rail get subsidies, even in Europe and Japan - it just isn't as much as in the USofA.

But airlines get some subsidies. How much, and for which routes? that I do not know, and am unsure of where to find such data. Until we can get a number of "X dollars to move 1 person between Y and Z" that includes everything, we (the people) will just be pissing in the wind.

Forgive me if I presume Musk is looking for a hand out. Because I'm pretty sure he is.

> It's a grand idea and I'd ride
> one for fun.

It's not an "idea." It's a daydream. It's the opening shot of a science-fiction TV show, before you show the creatures from other planets or spaceships beginning an attack.

> Well, it won't be nearly as
> expensive

What's the atmosphere like on Planet Credula? Is it full of hydrogen and oxygen, like our own here on Earth? Does it have an exhilarating scent after a cleansing Springtime shower?


Guys… I beg you… Read the Levy link in the first comment. This train is NOT real. It's not real-adjacent. It doesn't even send Christmas cards to real.

so what happens when you are going 700+ mph, and I dunno... you get a 4.6 richter tremblor that move one of your in-the-sky pylons a few inches?

If the tube remains connected? prolly nothing. if it moves at a hardpoint where the tube disconnectes?

Well, trying to figure which bit used to belong to which person is daunting. And the whole enterprise is down until or if it gets fixed after the funerals and investigations.

Honestly? The people pushing and interested in moving this way, could easily all work from home, and LET THE ELECTRONS MOVE.

Business travelers already HAVE a way to do this, that already has infrastructure, and conveniently is available to tourists too. It's the most modern thing evar!!1111!! and it's called Jet-Air-Travel.

Vacuum tube travel sounds cool, but it is MARGINAL, ie. it is only marginally better than what exists. The premium to make it exist too, takes away any cost advantage.

And? The environmental impact statements alone for vacuum tunnels will take the better part of 30 years, so, 'don spect this in your lifetime.

The idea that this will be faster and easier 'cuz no security screenings/airport headaches is an illusion, becasue it will take only one incident for the TSA to have to take over the security "because: govt. reasons"

It's a pig-in-a-poke that the average person would never use, but they will be forced to slave away for.

"But airlines get some subsidies. How much, and for which routes?"

Look up "Essential Air Service". It's mostly subsidies for airlines that maintain flights to third-tier destinations. However, since 1990, it has been scaled way back, e.g., the government used to pay Delta to do a flight each day to Cookeville, Tennessee. It no longer does, so Cookeville no longer has any scheduled airline flights.

Remaining subsidies are mostly indirect, in the form of the air traffic control system that the FAA pays for, and airports that get FAA grants for improvements. However, these are funded almost entirely out of ticket taxes, fuel taxes, and landing fees. They take little revenue from the general fund. The FAA's budget from the Treasury goes mainly towards regulatory and enforcement, with some towards R&D.

There is a piece of NASA's budget that goes towards aviation research also. In federal-government terms it's pocket change, $1B or so per year. Most of it is far-future stuff, like hypersonic aircraft and scramjets.

"Anyone should notice that trains have to stop to get people on and off them. "

I've actually seen a concept for a clever solution for this: a train with a self-powered "drop car" at the end. When the train approaches a station, people who are getting off at that station move to the drop car. Near the station, the drop car uncouples and then gets switched to the station siding, and eventually pulls up to the platform and stops. Meanwhile, another drop car has already departed the station with passengers who are boarding. It merges into the main line right behind the train, pulls up ot the end, and then couples itself. Passengers then move from the drop car to the main body of the train, and then the drop car is ready for the next stop.

How fast the train can go depends on how fast passengers can be rotated in and out of the drop cars, and how far apart the stations are. The technical challenges do not appear to be huge; most of it is existing high-speed train technology. There are logistical considerations, e.g., what do arriving pax do for local transportation? (One solution: put the stations at airports and let them rent cars from there.) The economics is obviously the elephant in the room, as it is for any high-speed train service.

"Passengers then move from the drop car to the main body of the train, and then the drop car is ready for the next stop."

Any system that relies on passengers being uniformly alert, mobile, and punctual in moving* to,* and being* in * the correct cars is doomed to failure.

And comfortable assigned seating with plug ins for your electronic devices like they have in Japan on the bullet trains is a non starter.

Driverless cars will kill the whole scheme.

"Driverless cars will kill the whole scheme."

I worked on engineering studies for auto-piloting cars back in my graduate-student days in the early '90s. My take: there's only one way it will ever happen in the U.S. Our liability regime is just too onerous; it makes the business risk essentially infinite. The one way it might happen would be if the federal government takes over the system so that it can exercise restrictions on when and where people can travel.

1. Insurance companies will see that DL cars are safer.

2. Commuters will see that they're faster.

Each will make a few calls.

Done.

Okay, so I didn't realize that Aggie at December 22, 2014 8:16 AM was being purely sarcastic, which he was, so my snark is withdrawn.

(Yet still within reach! Trust NO ONE!!!11!!)

(Note Swiss-style insincere emphasis. See? I play well with others!)

worked on engineering studies for auto-piloting cars back in my graduate-student days in the early '90s. My take: there's only one way it will ever happen in the U.S. Our liability regime is just too onerous; it makes the business risk essentially infinite. The one way it might happen would be if the federal government takes over the system so that it can exercise restrictions on when and where people can travel.
Posted by: Cousin Dave at December 22, 2014 2:23 PM

I'm sure you are a great engineer, but I don't think you understand that liability is created, and limited by the legislature, and the courts.

The federal government and state government can easily limit liability and risk, (just as they do for medicine, aviation, and driver directed cars, just by passing a few laws.

If medical liability was allowed to run amok doctors and hospitals could not stay in business.

When you look at comparative risk, human drivers are horribly risky, they fall asleep, they get distracted, and they make mistakes.

If your theory was correct airlines would not have computers controlling planes. The liability would be a problem.

The risks of a computer driving your car are infinitesimally smaller than a real person. Auto insurance and legal liability can and will be structured to take that into account.

I remember when Price-Anderson survived a challenge in 1978. The PR teams for the energy company behind the doomed Marble Hill project flatly lied about it... Saying in as many words that the Act compelled nuke operators to insure themselves for all expenses after an accident.

"The federal government and state government can easily limit liability and risk."

But when does it trouble itself to do so? Very rarely, from what I've seen. Computer-controlled cars may in fact be safer... but it only takes one malfunction, one big pileup, to bankrupt everyone involved with legal costs. And as we often see in aviation, this extends to the "neighborhood". Boeing had to spend millions on legal costs after 9/11, even though the federal government partly indemnified the industry. This is despite the fact that any sane person would conclude that Boeing had nothing whatsoever to do with the planning or execution of a terrorist attack. The judiciary has made it very clear that it isn't going to lift a finger to help solve the problem... all it would take would be some groundless lawsuits being immediately tossed out after a show-cause hearing, and a few of the purveyors of the most abusive actions being found in contempt for bringing such garbage to the court, but it ain't gonna happen, anywhere, ever.

Cessna survives today in the U.S. because of the Kassabaum Act, which placed a statute of limitations (17 years IIRC) on liability having to do with general aviation products. Had it not been for that, today Cessna would be either offshore or out of business, along with many suppliers. We know this because it came within an eyelash of happening before the law was passed. Will Congress ever enact reasonable liability limits for other industries? I've been around some product liability lawyers, and without exception they've all been amoral cutthroats, concerned with nothing other than how much money they can make by issuing a steady stream of threats. And the worst part is that they've all convinced themselves that they are saints doing God's work. They are extremely wealthy and have deep political connections. Ordinary people like us cross them at our peril.

Getting back to my point, the technology to implement driverless cars has existed since about 1995. Why hasn't it been done already? Liability. There's simply no way to project or contain the costs. In the event of a big crash, liability and legal fees could run into the hundreds of billions. No investors will take what they perceive to be an infinite risk. No insurance underwriter can write a policy big enough.

Well, we have big pileups all the time… Thousands die each year, but manufacturers keep making cars and people still keep buying them. This will certainly involve some new law, but Google's been lawyering up in the Beltway, and DC must be pretty horny for the seduction.

I see what you're getting at, but this is obviously going to happen... It's about where we were with portable phones in, say, 1990. Law and regulation are out of control, but Google is a huge company with brilliant people and a lot of resources, and I bet they'll be ready to power through.

Let me put it this way: Our Godforsaken government is a lot better at assembling huge new volumes of regulation than it is at building technically dicey vacuum train-tubes on stilts along world-famous continental transform fault lines.

We'll see. From a legal standpoint, everything's going to change once the automation is introduced. So I'm not convinced that anything having to do with they way actions over past pileups were handled will be viewed as a precedent.

Well, we have big pileups all the time… Thousands die each year, but manufacturers keep making cars and people still keep buying them.

One thing that keeps the liability down on cars is that the operator is an individual. if the car crashes due to operator failure, it's just an average guy. Can't sue the manufacturer over operator error.

If the train crashes due to operator error, sue well-funded operating company.

Driver-less cars mean the operator is Google, not Joe Sixpack. Big wallet.

> Google, not Joe Sixpack.
> Big wallet.

Right, so we'd expect Price-Anderson style indemnity. Google will ask for it, and since taxpayers want the innovation to happen despite risks (as with nuclear), they'll probably get it.

> I'm not convinced that anything
> having to do with they way actions
> over past pileups were handled
> will be viewed as a precedent.

Well, we're not going roll civilization back to the Code of Hammurabi and take it from the top... Something will be used as precedent. You guys seem to have very specific ideas about how progress will be congested in this case.

For innovations of this magnitude —changes which involve international competitive leadership and prestige, logistical refinement throughout the industrial and social spheres, and consumer demand— I think we can and should expect human desire to find a way.

'Sides, we got nowhere else to go. Follow the Levy link: Hyperloop isn't real.

Let no one be distracted from my climbdown of sarcasm towards Darth on December 22, 2014 4:03 PM: I was wrong, and he was properly cynical.

Driver-less cars mean the operator is Google, not Joe Sixpack. Big wallet.

Posted by: Conan the Grammarian at December 23, 2014 9:12 AM

Only in cases where the crash is caused by a malfunction of the on board computer.

(Most accidents will not be )

GM is on the hook for mechanical failure that causes injury or death. But getting to those deep pockets requires causation.

The computer will create a record of all events leading to a crash.

Insurance premiums will go down drastically for computer operated vehicles.

The biggest problems are going to occur when the traffic is mixed between human operators, and computer driven vehicles,

As usual, poor drivers are going to attempt to point the finger at a machine for causing the accident.

As usual, poor drivers are going to attempt to point the finger at a machine for causing the accident. ~ Posted by: Isab at December 23, 2014 4:18 PM

If they're in a self-driving car...?

As usual, poor drivers are going to attempt to point the finger at a machine for causing the accident. ~ Posted by: Isab at December 23, 2014 4:18 PM
If they're in a self-driving car...?

Posted by: Conan the Grammarian at December 24, 2014 11:06 AM

No,the ones who are still riding around in manually operated cars. Isn't going to happen overnight you know.
There will be a significant number of non computer driven cars on the road for the next thirty years.

Although I suspect that the degree of computer controls in newer model cars today will mean that there will be a substantial demand for retrofitting.