Live data from Hacker News

Netflix and Alphabet will need to become ISPs, fast

techcrunch.com

141–150 of 284 posts

Re: Netflix and Alphabet will need to become ISPs, fast

#141

I don't understand why this is turning into such a massive panic. The internet wasn't cut up into throttled tiers pre-NN, and who exactly is going to be paying for internet if it doesn't deliver what they want? Why do people have so little faith in the market? Why is the answer MORE regulation, when regulation and regulatory capture are what got us into this monopolized mess in the first place? What about the fact th…

1. Monopolization has increased in markets in general (see: ATT-Time Warner merger). 2. ISPs have realized what they can do with their increased monopolies, and have started developing technical strategies for doing so. 3. Regardless of this, these companies are getting certain things from the public, such as right-of-way. They act like they owe nothing to the public, which is a lie. It's a logical mistake to assume…

Re:

1. ATT and Time Warner didn't really compete prior to the merge.

2. ISPs have realized that basically every business is dependent on them to access their customers. Why wouldn't you try to extract value from that?

3. Hand-wavy feelings doesn't mean anything in government outside of impassioned speeches. They owe the public nothing other than providing their service according to the contracts they made. Large organizations where every individual has their own incentives don't really have feelings or a sense of duty.

---

1. This is already true, as it turns out private citizens largely don't actually own the land where the poles are and governments don't want to charge.

2. You would hate this world. Content can't be encrypted because there might be illegal content contained within. All your traffic will need to be scanned and scrutinized. P2P is basically dead. Any site with user-generated content will be blocked. Sites will have to make legal agreements with ISPs to be whitelisted for strict content moderation.

3. Blocking municipal ISPs is not as mustache-twirlingly evil as people make it seem. Some ISP goes to a town where it would normally not be profitable to operate and makes a deal with city in exchange for exclusivity. The risk that the citizens of that town would just form a municipal ISP funded by tax dollars and undercut them once the market is established has a real cost which is going to be paid in one way or another.

Re: Netflix and Alphabet will need to become ISPs, fast

#142
post #47
post #40

Earlier quoted context omitted.

You must live in one of the few places in the US where high-speed internet is available through multiple providers. That is definitely not the norm. Most people won't be able to simply switch to the least nasty provider.

If the FCC really wanted to promote competition instead of profits, they could mandate nationwide access to utility poles & shared conduit. As well as making exclusive franchise agreements illegal.

Exclusive franchise agreements are illegal for eg cable. They have been illegal for a quarter of a century.

You'll find that nearly all examples commonly presented of the lack of competition, are not due to market lock-out, it's due to the high cost and time required to gain approval to build in a given town or city. The combined cost, with pre-existing competition entrenched, is a very large disincentive. Google Fiber didn't grind to a halt on their build-out because they couldn't legally access markets, it's because it's painfully expensive and very slow.

There are three solutions to the problem: new technology that bypasses (low earth orbit satellite, wireless); lower the total cost to enter markets and compete; mandate shared access to existing infrastructure (and ideally simultaneously lower the other market entry costs).

Re: Netflix and Alphabet will need to become ISPs, fast

#143
post #128

Since I work for a startup ISP I am praying that AT&T will be dumb enough to prioritize their content over others. You're going to make TNT fast and Netflix slow? You've got to be kidding. The amazing PR team at Netflix will be all over this and users will revolt. It will open up opportunities for the small guys to compete. By all means big ISPs, please cripple a huge part of your service that is very important to yo…

I think this comment hits the nail on the head

- It will open up opportunities for the small guys to compete.

Re: Netflix and Alphabet will need to become ISPs, fast

#144

Earlier quoted context omitted.

More than half of households have two or more broadband providers at 25 mbps+; with 5G, the typical household is going to have 2-4 such providers. Two providers isn't a highly competitive market, but it's the same level of competition as exists for mobile platforms, and that seems to work fine.

Sorry, you're looking at that wrong - almost half of US households only have a single choice for ISP. That's not competition, it's a travesty.

Only if you exclude wireless options, which is becoming an ever-less rational thing to do. (15% of households making over $100k/year have ditched wired broadband in favor of wireless, and that number is growing rapidly.)

Also, the question is which households have options? I strongly suspect its the households in wealthier metro areas and suburbs. Competition for households that are the largest sources of potential revenue is going to drive the market. It's exactly like mobile devices--a huge fraction of the market has just one choice of smart phone platform (Android), because Apple doesn't sell cheap phones. But competition between Android and iOS in the mid-range and high-end drive the behavior of the market.

Re: Netflix and Alphabet will need to become ISPs, fast

#145

Earlier quoted context omitted.

This type of comment implies you think the world is static. I'm not saying that this good, but what makes you so sure market and social forces won't produce a better model? To me this is just a temporary setback, and a reminder why we need to focus on developing new technologies, like a "truly decentralized internet", blockchain, cryptography, and more. It almost feels like a the death throws of a monster that will d…

None of those technologies matter when your access to them is blocked. When it comes to access, you have copper, fiber, and radio. The latter is not really competitive to the first two, both of which require significant up-front investments with no promise of revenue to pay those investments back. Remember Google Fiber? Where's that now? Shut down with a marketing-speak comment that says "its purpose was fulfilled."

They might not be expanding, but I wouldn't call it shut down. I have an active account and they're providing my internet service in KC. This was set up this past February.

Re: Netflix and Alphabet will need to become ISPs, fast

#146
post #37

Earlier quoted context omitted.

> 2) Google will need to lobby for changes in local regulations that prohibit local ISPs... which helps the local ISP startup market. Google won't do that because of its political ties. To open up local ISP competition, you need to get rid of things like build-out requirements, loosen permitting requirements, reduce the ways people can stall installation of Fiber huts for NIMBY reasons, etc. Google will happily push…

That sounds exactly backwards. Google argued for net neutrality because they said it benefits everyone, not just them. And it does benefit everyone. Google doesn't have to justify things politically. At the same time, Google doesn't always do everything it can in the public interest, it sometimes does things that primarily benefit itself.

Net neutrality is something that Google can reconcile with its other political affiliations. The stuff that's required to reduce barriers to broadband deployment at the state and local level is not. Take build out requirements for example (the idea that ISPs should be forced to build service to areas where people can't afford to subscribe to it). Google can't lobby to get rid of build-out requirements generally, even though it would massively help broadband deployment. It's a meat-and-potatoes Democratic issue at the state and local level. So instead of lobbying to get rid of them, it simply sought waivers of those requirements in every Fiber city. (And even then, it got pilloried for it: https://www.huffingtonpost.com/alex-salkever-/is-google-fibe...).

Re: Netflix and Alphabet will need to become ISPs, fast

#147
post #91

Earlier quoted context omitted.

That means nearly half have a single (or no) option. Further, for many who have two choices, the decision is between not-actually-very-Fast DSL or a single cable provider. In Mountain View I had only shitty DSL or shitty cable. Both sucked and cable was very expensive. (This may be different now. My experience was 8 or so years ago.) I happen to be lucky to have actual high speed Comcast now, with an option for FTTH…

>for many who have two choices, the decision is between not-actually-very-Fast DSL or a single cable provider Very true. In one city where I have a house, the choice is 4 megabit DSL for $60/month, or 300 megabit cable for $101. Both are considered "broadband" by the FCC because the DSL is "up to" 8 megabits. But they are not equivalent services. There's also a wireless ISP, but it's $95/month for 5 megabits.

The FCC definition of broadband is 25mbps, not 8mbps.

There was a proposal to roll that back, however it didn't succeed.

You'll find that companies like Verizon do not call their slow DSL services "broadband." In Verizon's case they say: "High Speed Internet (DSL) service."

https://techcrunch.com/2018/01/18/fcc-report-keeps-faster-de...

If you look at Verizon's DSL page:

https://www.verizon.com/info/dsl-services/

You'll see they can't call their actual service broadband if it's not up at 25mbps. Instead they throw the "broadband" term around on other things on the page, such as "broadband routers" and they call themselves a "broadband provider" (referring to their FiOS offering).

Re: Netflix and Alphabet will need to become ISPs, fast

#148

Earlier quoted context omitted.

This type of comment implies you think the world is static. I'm not saying that this good, but what makes you so sure market and social forces won't produce a better model? To me this is just a temporary setback, and a reminder why we need to focus on developing new technologies, like a "truly decentralized internet", blockchain, cryptography, and more. It almost feels like a the death throws of a monster that will d…

> This type of comment implies you think the world is static. I'm not saying that this good, but what makes you so sure market and social forces won't produce a better model? They already are: multiple low or modest cost, high speed satellite providers are targeting the US market for deployment in the next four or five years. The dumber Comcast and AT&T are about the situation, the more money SpaceX is going to make…

> They already are: multiple low or modest cost, high speed satellite providers are targeting the US market for deployment in the next four or five years.

Hoping satellite will solve your bandwidth problems is wishful thinking of the worst sort. There simply isn't enough spectrum, and thus bandwidth, to go around.

Can't beat physics.

Re: Netflix and Alphabet will need to become ISPs, fast

#149

Another strategy to this (not exclusive of other strategies) is to encourage municipal ISPs, or even go further and roll out federal ISP infrastructure ala the federal highway system. I've always thought that this would be well within the rubric of the USPS. The constitutional mandate for a postal service was drafted at a time when the internet could not have been conceived of. I would argue that the intent of the co…

The next step beyond the mail was the telegraph and then the telephone. Those were built out by private companies, but operated under a regulatory regime called "common carrier," which protected the network owner from liability so long as it charged and carried traffic from all parties equally. These days, that concept is called "net neutrality" in the context of the Internet, and I suppose you may have heard it ment…

>A government-run economy becomes starved for capital and innovation slows way down; that is why the Soviet Union lagged U.S. technology and why Cubans still drive 50-year-old-cars.

This assertion of causality is sorely lacking in accuracy. Your examples alone are sufficient for highlighting this.

Being among the most ‘government-run economy’ on earth at the time, the astounding innovation achieved within the Soviet Union is the most glaring contradiction. Cuba is just an awful case study subject, due to so many drastic incongruencies, but let’s remember technological innovation has been fairly low on Cuba’s priority list. If it weren’t their access to the world’s selection of material and media resources would have likely remained open. And this access was very obviously the reason they are driving old cars. Those are just two factors I will assert are far more useful in this sort of prediction.

While of course the correlation is circumstantially critical to varying degrees, many other measurements are better predictors of rich innovation environments.

Re: Netflix and Alphabet will need to become ISPs, fast

#150

Another strategy to this (not exclusive of other strategies) is to encourage municipal ISPs, or even go further and roll out federal ISP infrastructure ala the federal highway system. I've always thought that this would be well within the rubric of the USPS. The constitutional mandate for a postal service was drafted at a time when the internet could not have been conceived of. I would argue that the intent of the co…

The next step beyond the mail was the telegraph and then the telephone. Those were built out by private companies, but operated under a regulatory regime called "common carrier," which protected the network owner from liability so long as it charged and carried traffic from all parties equally. These days, that concept is called "net neutrality" in the context of the Internet, and I suppose you may have heard it ment…

I'd disagree. Government is much more efficient at funding capital projects. Public companies are largely incapable of sustaining them at scale.

The Bell System was built through implicit subsidy of capital expenses, by the allowance of monopoly, favorable tax treatments of convertible bonds, and a regulatory regime that locked in profit.

Post reply on HN