Google is rolling out fiber to neuter Comcast. Period.
Comcast basically blackmailed Google several years back and thought they won. Google, rightfully, now saw Comcast as an existential threat to their business.
Google will keep rolling out until they have Comcast by the balls. Only then will they stop.
If Google wanted to come at Comcast, they'd be gearing up a lot faster than they are. At this point Google Fiber is mostly FttPR-Fiber to the Press Release. And also, I don't think any of the first several Google Fiber locations were even in Comcast territory. So far AT&T and TWC have been their competitors.
They want to create demand because Google loves remote computing and loves internet consumption.
Google is providing a model for the ISPs by forcing local communities to deregulate their ISP franchising system. Before Google Fiber, each town or state would take a pound of flesh from the new company trying to lay down a network. Google made the cities bend over backwards.
https://gigaom.com/2010/02/10/google-doesnt-want-to-be-an-is...
Our cap was 50 Mbit which wasn't bad, but then when Google came in, Comcast upped that plan to 250 Mbit with no hardware upgrades or anything for the same price which was really nice.
The main thing I noticed though is that reliability with Google is a lot higher. My Comcast connection would go down for an hour or two about once a week, but my fiber connection seems to always be fine.
As far as cities bending over backwards, Provo definitely did that. We had put in a citywide fiber network using a $40 million bond and then ended up selling it to Google for $1. It was definitely the best move considering the company that had been operating it before defaulted on their payments, but Provo will still be paying for the network for 12 years.
All of their products are designed to give them different kinds of data, and being the man-in-the-middle to the entire internet on a per-household basis gives Google an incredibly fine-grained picture of the patter-of-life of the people that use their ISP service.
Being able to steal Comcast's market is just a nice side benefit.
See also Google Fiber's privacy policy:
Other information from the use of Google Fiber Internet (such as URLs of websites visited or content of communications) will not be associated with the Google Account you use for Fiber, except with your consent or to meet any applicable law, regulation, legal process or enforceable governmental request.
https://www.eff.org/deeplinks/2014/11/verizon-x-uidh
A variety of ISP's (Comcast, MTNL, IDEA) also run MITM attacks and inject ads in pages, sometimes on behalf of law enforcement. I imagine they inject tracking as well.
https://blog.ryankearney.com/2013/01/comcast-caught-intercep...
https://medium.com/@sushubh/delhi-police-is-literally-doing-...
Google could either do these things or block others from doing them.
Honestly even if you take that as meaning they won't ever track history of internet access to a google account, why would you assume they need to do that, just to have more information on you?
It's not like you escape their privacy invating antics if you just use google search (or a site with google ads) without an account.
They get some of this already when people choose to hand Google their DNS traffic, but being the ISP makes this complete. You may want to also watch Aral Balkan's talk posted just above DJB's at the same URL; he does a better job than I can at describing how Google (and Facebook, etc) are using surveillance-as-a-business-model.
> Google Fiber's privacy policy
The risk of a national security letter (which the part you quoted includes as an exception) is already enough of a reason to limit how much personal data is in a single database, that's not the main problem. That policy isn't a binding contract. Do you really think it won't be changed the moment someone at Google decides it would be profitable?
They promised "Nest" data would stay separate, too, which quickly changed to sharing "limited" data. Sorry, but promises not to snoop don't mean much when the promise contradicts their business model.
[1] https://projectbullrun.org/surveillance/2015/video-2015.html...
Your argument doesn't make sense, especially not with the rapid migration of traffic to SSL (and no, Google aren't MITM'ing SSL connections). The lifetime value of a single family's worth of ad-eyeballs simply isn't there to justify an undertaking of the size of Fiber.
Yes, they are ultimately an ad-company, but that doesn't mean that there is always a single-hop connection between any Google activity and ad revenue. If there's an ad angle here, it's in enabling end users to consume much larger quantities of ad-supported content.
When 95% of your revenue comes from ad sales, you're an ad company.
Perhaps it makes more sense that Google is investing in high-cost infrastructure because they want to move away from the ad-model business. It's risky to have 95% of your revenue coming from a single industry.
Google can eat a lot of the costs of setting up a better network, which is a great investment in the future if you're betting that a majority of people will need high speed internet access 10-20 years from now. Then they can start extracting monopoly profits just as AT&T, Verizon, Comcast etc. do now.
Google has many skills, but last-mile delivery networks aren't one of them. Not that they can't be, but it's just not that profitable of a business to waste their time on. The business itself is just a math problem balancing quality of service with investment -- and not a particularly difficult one.
Fortunately, that's not their play here: Google has invested far too little money in Fiber to be truly serious about it as a business. Fiber is just their way of generating consumer demand for higher speed pipes by saying "hey, look at this! this is how it should be!" They did the same thing on Android with the Nexus phones, so it's not even a new tactic for them.
> Google can eat a lot of the costs of setting up a better network, which is a great investment in the future if you're betting that a majority of people will need high speed internet access 10-20 years from now.
The problem is that you can't just build the network and be done with it. It costs a lot of money to maintain and provide service (approx. 10% of the original cost of the network per year), and that's money that Google will want to spend on other, more profitable things. Hell, if they want to get into infrastructure, the smart play would be free wireless Internet for the developing world (i.e. what Facebook is doing).
But again, Google hasn't spent anywhere near the amount of money on Fiber that they would need to in order to make it a commercially successful business. Because that's not the point: they're using cities like Kansas City and Austin to showroom their vision of what the Internet should be.
Google will never expand Fiber to the entire country, and I'd be surprised if they don't sell the infrastructure to another company in 5-10 years once they've achieved their goals. If the ISPs are providing better service, customers can upload more personal / sensor data to the cloud for Google to mine, and Google can cram more ad-laden services down the pipes. It's a win/win for them.
This whole endeavor is about Google flexing its muscles as a media power player to ensure their content/ads aren't held for ransom by any ISPs. Any company can try to build a national fiber network; but it takes Google (and their army of media contacts thanks to Google News) to talk this much about what amounts to a minuscule infrastructure investment (~$150 million over 5 years - or about what Comcast spends on infrastructure in a week).