Private infrastructure | Hacker News Reader

Private infrastructure

aaronkharris.com

37 pointskatm20 comments

I think the general point to take away from notions like this is that many of these very profitable opportunities (Uber, airbnb, instacart, even private MOOC to replace education) are all responses to public infrastructure failures as our voting population becomes more and more self-centered and focused on individualist (i.e. extremist capitalist/Libertarian) economics and less willing to fund public infrastructure, be it in transportation, education, any public good really.

The solution that is always suggested is some form of leaning in harder to Libertarianism and trying to explain away how private control of infrastructure with a profit motive is somehow a good thing. Which makes sense taking the world view that [private profits == good, market economics are the only solution to any problem] for granted and if you're likely to be the private person controlling the infrastructure and its corresponding enormous untouchable profits. Incidentally, the private structures are even certainly better for everyone than the broken dilapidated infrastructure that exists right now (see: Uber) and so it's a net gain for the public in the short term, but it's dystopian in the long term.

> these very profitable opportunities (Uber, airbnb, instacart, even private MOOC to replace education) are all responses to public infrastructure failures as our voting population becomes more and more self-centered and focused on individualist (i.e. extremist capitalist/Libertarian) economics

Not exactly. Yes, we do have a problem with voters playing identity politics and voting tribally -- to promote one constituency over another. Overwhelmingly, these are not libertarians or fans of capitalism. Elected representatives bring home boondoggle infrastructure projects with the primary intent of "creating jobs" and "getting reelected".

The problem is our representative voting model, which has been largely unsuccessful in the last 50 years at maintaining or building out new infrastructure. This model is very good at rewarding the in-group and punishing the out-group, often by a winner-take-all tyranny of the majority, but it's ineffective and inefficient at actually getting things done or satisfying the voting public overall.

Governments are huge today compared to the past, and still growing. This suggests that a focus on individualism is not among the reasons for any decay of infrastructure compared to the past.

Governments have expanded in order to better serve private wealthy interests. They have expanded in response to individuals pushing for "what's good for me" and ignoring "what's good for the public/country". Hence individualism.

The public infrastructure which you've described didn't 'fail' because of lack of funding, it 'failed' because of lack of aligned incentives. Please specifically outline how Uber and similar services are good for only the 'short term' and will be 'dystopian' in the long, I'm very curious.

Of course it did -- public transportation, the alternative to mass widespread use of taxis/private cabs most certainly dies in most of the US because of a lack of funding.

And you're right I left that dystopian comment vague -- let me clarify, having a large private monopoly own a necessary piece of transportation infrastructure is my idea of a dystopia, incidentally especially when their business model is built around using ownership of the name and app to exploit people who really need to use it (i.e. surge pricing) and people who work to provide it (i.e. the "technically private contractors but really just employees without benefits").

The point is transportation is something we all require and should be provided by the public (it already is in the form of roads, but that's not good enough): taxis are a crappy but elastic solution that's there to fill in the gaps. I think their awfulness is a good thing: private cars driving people around are a terrible solution to the real problem which is transportation infrastructure. As soon as a terrible solution starts to look and feel okay to a lot of (wealthy or middle-class) people, that's dangerous. Because things like Uber don't look and feel okay to the very poor. That's why I describe it as dystopian.

Another self-serving Libertarian-themed post that works under the flawed premise that market forces can replace all natural monopolies efficiently, as well as surpass the laws of physics.

There's a whole bunch of reasons why you don't want infrastructure to be privatized, the biggest of which being because infrastructure will not be profitable everywhere, yet that doesn't take away its strategic value not the idea that there are some fundamental services that should be reasonably available to all citizens.

I didn't seem to read the same vulgar libertarian subtexts that you did. It was more an outline of computer technology enabling an easier proliferation of alternatives to state-provided services. There isn't much of a value judgment being made, in fact it notes that "opportunities are being exploited by a range of startups and companies who are essentially building private infrastructure, mainly for the relatively wealthy".

I see the absence of a value judgement in the observation of the infrastructure of a developed country falling apart and as explicitly saying that it doesn't really concern him.

As a citizen, investor, whatever, that fact should be frightening for all sorts of reasons. The investment opportunities are, if anything, a minor consolation.

I don't think you'll find a lot of people cheering for obscenely bad causes, but you'll definitely find a lot of people who have very neutral opinions and opportunistically seize business opportunities without much regard of the externalities and systemic effects.

Which critical infrastructure cannot be profitable?

I have never seen buses and trains being profitable.

Postal services in rural areas operate at a loss. Same thing with internet and phone connectivity.

I'm very skeptical on all three, do you have any supportive data?

It's a Google search away. Seriously, the story of the cost of supporting public transportation. But you can have a link to the latter assertion http://www.outsidethebeltway.com/should-we-continue-to-subsi...

It is most certainly not. The inefficiency of USPS doesn't speak to the inherent inability for buses, trains, postal services and internet providers need to be subsidised. That it a huge claim you've made, one link on the inefficiency of USPS(a government funded service...) does not suffice.

The inefficiency of the USPS speaks to the inefficiency of government services. They've lost $47bn over 10 years. Why? Because they have no incentive to care about their business model, because they can be apparently infinitely supplied by taxpayers.

I don't really see how Instacart can be considered "infrastructure", at best it's a marketplace-driven service. There are lots of examples of private infrastructure - e.g. the internet - which don't suffer the various problems described, and all the examples cited still rely on actual infrastructure to exist..

It's a layer of abstraction. Like Uber or APIs or development frameworks like Rails. The layer of abstraction allows other people, services, companies and ideas to be built on top of them.

For example, Instacart and Uber allow neighborhoods that aren't near grocery stores and transportation hubs to be more viable because they can replace the functionality of needing to have those near by.

I don't disagree with that, but it doesn't really answer the question. "Infrastructure" doesn't mean the same thing as "layer of abstraction", there's an essentialist quality to infrastructure.

You could build an infrastructure _using_ self-driving cars (as an example), but they are not an infrastructure in and of themselves. There's not a huge heap of difference between Instacart and bike couriers (except they provide slightly different services), and while couriers are a very useful service they're really not infrastructure.

I don't think Instacart is, or ever wanted to be an infrastructure. At least not as a transportation infrastructure. There is an argument to be made that it is becoming an infrastructure "within" a supermarket, mainly because of their pick & pack infrastructure (this is probably what Aaron is referring to) and understanding of inventory. Compare that to what we're doing at Postmates. We're in the process of creating a transportation layer by combining tens of thousands of people with cars and bikes or on foot into a "programmable layer" that can be accessed through our API and utilized by third parties. Both have huge potential.

It's infrastructure so far as it is a supplementary good to physical infrastructure. You could think of it as virtual infrastructure. By the time you are cooking dinner, there is no difference between having an accessible grocery store vs Instacart. Similarly, when you are at your destination, there is no difference between having taken the subway and having used Uber.

The food desert comment was interesting but I guess it depends on your definition of a food desert. I imagine if everyone is delivering their food then food deserts (as defined as regions with walkable grocery stores) would increase as there would be less grocery stores around (replaced by gigantic food warehouses).

Reply on news.ycombinator.com