Hacker News new | past | comments | ask | show | jobs | submit login

If the US made any sense, this would be a public works infrastructure project rather than private enterprise. The tower density and spectrum that will be needed for effective substitution of traditional wired internet services with wireless towers will be prohibitive for any reasonable company to simply invest their way into, and big teleco has proven historically to love charging more for the same now than competing to make the service better once it starts looking expensive to keep up.



> If the US made any sense, this would be a public works infrastructure project rather than private enterprise.

No thank you. My Verizon plan is expensive, but it works pretty damn well, unlike most of the public services here. My wife and I took Amtrak every day between Wilmington-Baltimore and Baltimore-DC for about two years. It was regularly late (often very late). Trains broke down in-between Wilmington and Baltimore regularly, and my wife would be stuck in the train for hours waiting for a replacement engine to arrive. Many of the tunnels along the route are decades past their design life.

Before that, I lived in Chicago, where ancient lead water pipes poison kids. Before that I lived in Atlanta, where ancient sewer systems dump untreated sewage into the river whenever it rains. I also lived in Wilmington, DE, where bus drivers would just randomly decide to end their route 15-30 minutes early. Now I live in D.C., and even though I'm within 0.3 miles of a metro station both at work and at home, I take Uber because Metro trains keep breaking down and/or catching on fire.

I'm not a libertarian nutjob, but I don't want the government running my cell service. Our roads are awful (I just came back from Munich), our transit is awful, etc. There are a tiny handful of competent public infrastructure organizations in the country (e.g. New York's water system, Metro North), but most are a disaster. Verizon may be evil, but unlike say WMATA it doesn't have to regularly shut down major sections of its network because it spent decades neglecting its infrastructure.


Amtrak does not work not because it's a public service. Many railways in Europe are public and the work great.


Many public services in Europe work great. But I'm talking about the US.


Amtrak being late wasn't entirely their fault, The railroads that owned the tracks often delayed the passenger service giving preference to their freight service. They got dinged for that not too long ago.


Wilmington to DC is northeast corridor, which is all owned by Amtrak. That's why delays are usually only tens of minutes rather than hours (except when equipment breaks down and the delay is hours).


> The tower density and spectrum that will be needed for effective substitution of traditional wired internet services with wireless towers will be prohibitive for any reasonable company to simply invest their way into

It will be prohibitive for the government too, so lets just stick to wires.


Let's unpack this. If this is done at the Federal level, ala the Interstate Highway system, it would never work. Wireless infrastructure is still incredibly immature compared to road technology. And the idea that the government would be able to spend the amount of money required to keep up to provide the same or better service than the private sector has never been proven, and IMHO unlikely. Look at how NASA has been run the last 30 years, and how it's been spending its money on their current launch system.

The problem is that the capital expenses are high in this industry, so the players naturally want to recoup their costs and focus on profits. Eventually technology will shift, demand will rise, and one of the telcos will take the plunge in an attempt to gain marketshare. Their competitors will react, and for a while it'll be golden for the users. Then the cycle will repeat.


let's unpack the unpack.

> Wireless infrastructure is immature and [the government can't spend].

The exact same thing could be said for the computer industry in the 40s and 50s, which was almost entirely funded by the government, precisely because only it could spend enough money for big risk long term stuff. Ditto the internet itself. Ditto the airplane. Therefore there is ample evidence that the government can spend, and spend big, on strategic stuff, and there is ample evidence that, over decades, it pays off bigtime.

> capital expenses are high, recoup etc

That's why you want the government in, because it has a much longer investment recoup horizon, and the government, unlike private networks, benefits from the general indirect economic growth advantages of good mobile infrastructure. That's precisely why it is the government that builds roads, often at a project-specific NPV loss, but at a huge NPV positive for the economy, which it taxes. In other words, only the government is able to capture the collateral positives of good infrastructure, giving it a higher incentive to spend more than private companies.

> competitive cycle rinse repeat etc

Quarterly targets are a strong disincentive to long term capital intensive projects. Most such projects undertaken by the private sector have, at a minimum, government concessions to allow restriction of competition, pricing guarantees, or outright government financial support. Notice the whole thing where municipalities cannot expand their internet service? There you go.

In general the whole dogma that the government is always useless at everything, is very pre-2008, ie. dangerously incorrect.


Not sure about the downvotes, unless it's because of the new HN motto where everyone downvotes things they disagree with.

<In general the whole dogma that the government is always useless at everything, is very pre-2008, ie. dangerously incorrect.>

You set up a good strawman. And the implication that since 2008 (what exactly happened there?) we're somehow more capable of having well run, long term, government projects in the US is not supported by much evidence.

Yes, government can have a longer recoup horizon; but it also has an extremely short political horizon. If something doesn't succeed on the first attempt, it becomes fodder for whichever party is in power and supported it.

And while government has had successes in the past, as the investment folks remind us, is no guarantee of future successes, especially in something as complicated as creating a continent-wide wireless infrastructure. Imagine the cost, the rent-seeking, the cronyism, the NIMBYism that will infect such a project.

Government is not an agency like DARPA that is free to investigate and develop technologies without concern for budgets and satisfying constituents. Look at how politicized the FCC is. Imagine handing it perhaps the largest communication system in the world, and expecting it to manage it. Keep it upgraded. Understand which technologies are worth something, and ones which are dead ends. Oh, and make sure Grandma in Broken Bow, Nebraska has as good a connection as someone in Berkeley or San Diego.

And back to your strawman; show me large Federally run government programs implemented since 2008 that are on budget, meeting original goals, and not having huge unintended side effects.


You don't even need a wholly public-only solution. Licensing out state-owned towers to third parties, in addition to providing baseline public service, would be adequate to allow third party companies to compete too.

You don't need to try to artificially restrict spectrum use, just make intentional obstruction illegal just like blocking a road is. Then if state technology calls too far behind due to budget cuts people move back to private solutions until the state can get its act together again.

Of course, the problems with US infrastructure run a lot deeper than that. I imagine in a more democratic country you would see the maintenance and upkeep of such infrastructure much easier to maintain.


If you want to throw government agencies around: The NSA seems to be doing a fine job keeping up with current technologies.


Possibly. Other than Snowden releases, I don't know much about the inner workings or capabilities of the NSA to judge them. But consider the budget that the NSA has to work with.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: