Live data from Hacker News

Trust-Busting as the Unsexy Answer to Google and Facebook

lareviewofbooks.org

71–80 of 184 posts

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#71
post #59

Earlier quoted context omitted.

> Someone one else would've bought YouTube Google is probably the only non-telco who owns significant infrastructure. So other than Google, the only other option for YouTube would have been a telco. Are you sure you'd want a telco owning YouTube? Can you imagine how awful it would be? Vimeo is owned by IAC, who then owned AskJeeves, TripAdvisor, HotWire, LendingTree and today owns Tinder, OkCupid, Match.com, The Dail…

>Google is probably the only non-telco who owns significant infrastructure. So other than Google, the only other option for YouTube would have been a telco What about Amazon or Microsoft?

They both have a lot of datacenters and private fiber between them but neither has anything like the ISP peering network that Google owns. Not even close.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#72
post #64

Earlier quoted context omitted.

> Someone one else would've bought YouTube Google is probably the only non-telco who owns significant infrastructure. So other than Google, the only other option for YouTube would have been a telco. Are you sure you'd want a telco owning YouTube? Can you imagine how awful it would be? Vimeo is owned by IAC, who then owned AskJeeves, TripAdvisor, HotWire, LendingTree and today owns Tinder, OkCupid, Match.com, The Dail…

Any of the above options would be better than the current reality, where Google is basically controlling video on the internet. They're long past their days of innocence.

Yeah, I'm sure you'd love Comcast owning YouTube.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#73

Earlier quoted context omitted.

I don't agree with this at all. If you break up youtube from google, I think people will treat those very separately - they serve very different purposes. Same with facebook and instagram.

People already treat them very separately, I'd be surprised if most of the public (not HN) even knows that Google owns YT.

I assume the GP meant "the people in charge of..."

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#74
post #68

The moat that each tech company owns is its network. Rather than trust-busting these companies up into their different businesses, I think it would promote more competition to impose network interoperability requirements once the network is entrenched for too long. A useful example is craigslist: a company with virtually no innovation for decades that still dominates the classified listings space. Other companies tha…

You don't need new regulation; you just need to get rid of existing regulation.

It is currently illegal to build third-party applications once the first party tells you to stop.

Craigslist actually was involved in the court case that cemented this: https://en.wikipedia.org/wiki/Craigslist_Inc._v._3Taps_Inc.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#75
post #74
post #68

The moat that each tech company owns is its network. Rather than trust-busting these companies up into their different businesses, I think it would promote more competition to impose network interoperability requirements once the network is entrenched for too long. A useful example is craigslist: a company with virtually no innovation for decades that still dominates the classified listings space. Other companies tha…

You don't need new regulation; you just need to get rid of existing regulation. It is currently illegal to build third-party applications once the first party tells you to stop. Craigslist actually was involved in the court case that cemented this: https://en.wikipedia.org/wiki/Craigslist_Inc._v._3Taps_Inc .

> It is currently illegal to build third-party applications once the first party tells you to stop.

No, it is illegal to access the data of a third-party with your own app if that third-party doesn't allow you to do so.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#76
post #3

Would trust-busting do any good? The main reason no practical alternatives to Facebook exist is the network effect. The main reason that competitors to Google don't do well is that people will just use the search engine that returns the most relevant results (unless your internet connection is censored). The only practical solution is regulating ethical behavior of these incomprehensibly powerful corporations.

The main reason no practical alternatives to Facebook exist is the network effect. Doesn't it creep you out that one company has a monopoly on the social graph? When we busted up Ma Bell, we basically put into place standards for interoperation. What if we had a law mandating social graph companies open themselves up for interoperation once they reach some critical threshold of size? Basically, the idea is that once…

> At home, I use Duck Duck Go. It gives good enough results but comes with a better reputation with regards to privacy.

The fact that you do so, with no additional effort needed, kinda breaks the entire trust/monopoly argument, doesn't it?

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#77
One complement/substitute for this approach is mandating open standards. If Facebook had to be part of the Fediverse or something like that. The standards could be managed by a third-party, something like the Federal Reserve for APIs. This way they can move to accommodate fast-moving changes and competition from companies in other countries.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#78
post #4

As talked about in this article, I think the main catalyst to increased anti-trust scrunity will be a mind shift away from the "consumer welfare" view of anti-trust. For the most part, big tech companies provide amazing service. I can't think of another industry that has a product as good as the tech companies of today have. Yes there may be some fakes on Amazon, and DDG in some use cases may be better than Google, b…

Acquisition by Google is the only reason YouTube exists. If they had to pay for their networking they could not have survived (Google owns its own fiber and doesn't pay anyone for transit). If you are taking a consumer benefit angle I don't see how you conclude that Google should have been prevented from buying YouTube.

That sounds plausible at first, but is inaccurate on multiple levels.

By early 2006, YouTube had substantial peering in place. It's easy to prove: they talked about their programs at NANOG (https://www.nanog.org/meetings/nanog37/presentations/bill.no...). Without those, they would have folded already. They were acquired by Google only at the end of the year.

Google doesn't own all of its fiber. Also, some ISPs have made public claims of being paid by Google for transit traffic. Google, but also ISPs, saves a lot of money through peering or the Google Global Cache. This is also why YT had peering in 2006: remember that traffic moves across multiple ISPs and it wasn't just YT's costs that were going to go up with their exponential growth.

Last but not least, consider that Netflix doesn't own fiber – or at least haven't announced any – and they haven't been sunk by network costs, thanks to peering and edge caching.

[I used to work at Google and knew at the time quite a bit about YT's infrastructure. The above is all public information.]

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#79

Earlier quoted context omitted.

The main reason no practical alternatives to Facebook exist is the network effect. Doesn't it creep you out that one company has a monopoly on the social graph? When we busted up Ma Bell, we basically put into place standards for interoperation. What if we had a law mandating social graph companies open themselves up for interoperation once they reach some critical threshold of size? Basically, the idea is that once…

> What if we had a law mandating social graph companies open themselves up for interoperation once they reach some critical threshold of size? If we open up Facebook's data to other companies, what makes you think those companies are willing and capable of securing the data and respecting user privacy? Facebook having all of that data is bad enough. Requiring Facebook to share it with other companies would not make t…

For one, where sharing of data is concerned, none of this is about forcing Facebook to give your data to some random third party that asks for it. It is about you as a user being able to demand from Facebook that they make your data (or some selection of it) accessible to some third party of your choice.

But also, and maybe more importantly, interoperation does not really necessarily require any sharing of data. Think the telephone network: All telephone network operators world wide do interoperate, which is why it doesn't matter where you are a customer, you still can call any other telephone on the planet. But that doesn't mean that those telcos somehow share data of their customers, apart from the particular calls that happen between customers of two different telcos.

Also, one way in which Facebook could be forced to share data would be for migration to a different service: If you think Facebook is abusing your data, you could move to a different provider, and Facebook would have to allow them to pull your data over and then delete it on their side--and all that without you losing contact with people who stay on Facebook thanks to enforced interoperability.

Re: Trust-Busting as the Unsexy Answer to Google and Facebook

#80
post #14

Earlier quoted context omitted.

The main reason for that is because they have been allowed to build a walled garden. Interoperable protocols would allow other companies to innovate on an equal playing field. Indeed if we follow the Bell example, the networked platform could be forced to be separate from those who profit off of it.

Can you name some successful interoperable protocols used by social networks that allow for meaningful innovation and updates? All I can think of is Agora* (which no one uses) and IRC (which hasn't added nontrivial features in a very, very long time).

Whatever protocol Facebook is using internally to interoperate between their thousands of servers, with some additional authentication and authorization layers put in?

I mean, not sure how sensible that would really be, but I think it's worth considering that Facebook is not some PHP script with a small local MySQL database behind it either, but rather a massive distributed system in itself, even if under the administrative control of a single entity.

Post reply on HN