A better comparison would be with telephony companies of the old days. We now have the Telecommunications Act [1] that does for telephony what the article is suggesting for social networks.
Quoting from Wikipedia:
> Since communications services exhibit network effects and positive externalities, new entrants would face barriers to entry if they could not interconnect their networks with those of the incumbent carriers. Thus, another key provision of the 1996 Act sets obligations for incumbent carriers and new entrants to interconnect their networks with one another, imposing additional requirements on the incumbents because they might desire to restrict competitive entry by denying such interconnection or by setting terms, conditions, and rates that could undermine the ability of the new entrants to compete.
So either social media companies should be classified as telecommunications companies (imho not far-fetched), or we need something similar to the Telecommunications Act but for social media.
Great idea, except for the fact that this didn't work at all. The 1996 Act encouraged lots of outside investment in telco competition, but all of that is gone now. Unlike in more civilized parts of the world [0], USA regulators have not fostered competition in telecom. That's why we have 2.5 national telcos.
Having followed the question in France pretty closely, most of the progress in both broadband and mobile telephony came from mainly one slightly less greedy ant later comer: Free Telecom. Before that, it was basically 3 operators with several trials for illegal agreements on prices. Once Free was granted a license, France entered the 21st century and prices fell. Unlimited plans starting appearing.
The difference in terms of policy is not huge: the only difference is a slight difference in terms of corruption and outrage. The process in France was clearly corrupt, but a few institutions still do their job and users made it a political issue.
Getting that mobile license was a short call and it changed everything.
I’m not sure we can conclude Europe’s broadband policy is better than America’s. The former optimised for consumers today over investment tomorrow. That’s hamstrung an array of projects across the Continent.
Instead of trying to legislate social media companies on their terms, this would help in creating 1) a standard of communication across different ones and 2) help in creating better social media platforms.
I strongly disagree, social graphs are contextual and it makes sense to have different graphs for different social networks. When I friend someone on Facebook I am not agreeing to share some other info with them that I produce on a totally different site. And finding friends is not the hard part of using a new site. The network effect comes from people actually using the site, not from finding your friends, so this wouldn't make social networks more competitive.
It's entirely possible to leave this choice to the user who can determine what information they want to expose to whom while stile maintaining compatible standards for the platforms themselves. This is essentially how the federated Mastodon network already works.
Or for example, facebook, instagram, whatsapp and their new dating service which is all interconnected but optional.
Actually the EU's GDPR comes close to this with its requirements for data portability between websites. In particular, it states:
"In exercising his or her right to data portability pursuant to paragraph 1, the data subject shall have the right to have the personal data transmitted directly from one controller to another, where technically feasible." [1]
Conceivably this could be interpreted by EU regulators as requiring Facebook to automatically send all your Facebook posts (on an ongoing basis) to friends on other social networks (or at least to a shadow account that you've registered on some other network, which is then connected to friends in the Fediverse). If the reverse flow of data were similarly mandated (especially based on anti-trust considerations), then your friends on Facebook would be able to see what you posted on your Fediverse page. The hard part would be demanding that Facebook make your friends' posts visible in your account on another network, since that requires portability of someone else's data.
There are apparently efforts being made by Facebook and other big companies to enable certain forms of data portability [2], but I worry that these are designed to distract regulators and frame the problem in a way that is compatible with the business models of the major providers.
I'm pretty familiar with FOAF, LinkedData, RDF and all of the rest of the Semantic Web standards. But saying that all of these can replace a system like facebook is a bit unintelligent. Facebook brings much more to the table that users need which semantic web has never been able to accomplish like near-instant response times, UX that even my grandma can use and 99% availability... just to name a few.
Asking ever day users to create a site (buy a DNS and all that) and create a FOAF file and host it is kind of ridiculous.
Facebook is a centralized monopoly I agree, but a bunch of standards isn't going to replace the stuff that Facebook built.
I believe the problem of centralization can be solved by looking at the problem differently: what if we continue using these giants for hosting our data (with all the amazing benefits that that brings) but we urge them to provide us with ways to control the way they show us this data. We need better ways to explore our "connection graph", not just a static feed of people we know and things they post.
For instance I'd like to be able to change the algorithm that generates my main feed. I want to grab the "political liberal" feed from github (as an open source example) and load it onto facebook; then grab the "climate change biased" feed and load that and test how that works. I want to change the variables of the algorithm, test it for a few days see how that works and change it to something else if I see it's only reporting fake news.
This to me seems the problems we need to face this decade when it comes to social networks; building p2p networks from scratch, or using other decentralized standards is cool but isn't inherently solving the problem imho. There's nothing inherently wrong with the social media giants if they prove to us they can show us different views of the graph without getting in the way.
> Asking ever day users to create a site (buy a DNS and all that) and create a FOAF file and host it is kind of ridiculous.
It could work like email, where your ISP provides you with a FOAF (or more generally Social Linked Data) store as part of your Internet access package. Or you could buy federated hosting from your preferred 'cloud' supplier.
> Asking ever day users to create a site (buy a DNS and all that) and create a FOAF file and host it is kind of ridiculous.
This is literally what the article on Two-Bit History says in its conclusion as to why FOAF failed; I'm not quite sure why you're describing it as "a bit unintelligent" when it's, you know, saying exactly what you are?
> what if we continue using these giants for hosting our data (with all the amazing benefits that that brings) but we urge them to provide us with ways to control the way they show us this data.
That's a great idea in theory, but in practice, how do you propose to do that? Facebook and Twitter don't want to let you change the algorithm that generates your main feed; they don't want to give you ways to control how they show you that data. The API Twitter provides to third-party clients is heavily restricted (and has been very deliberately made worse over the years), and the API that Facebook offers for those purposes is even worse. Neither system wants to let you control your social graph. The kind of access you're describing is access that is difficult to monetize -- unless they charge for that access directly, in which case it's just as dead in the water in practice as "just embed this FOAF XML block on your personal web page" is.
> There's nothing inherently wrong with the social media giants if they prove to us they can show us different views of the graph without getting in the way.
I think from a purely technical standpoint, you're probably right, but we don't live in a world where "purely technical standpoint" wins the day. This is not their business model. This will never be their business model.
Mastodon and other decentralized networks have their problems, but given the choice between "improve Mastodon" and "convince Facebook to open access," only one of them sounds feasible to me.
> I believe the problem of centralization can be solved by looking at the problem differently: what if we continue using these giants for hosting our data (with all the amazing benefits that that brings) but we urge them to provide us with ways to control the way they show us this data. We need better ways to explore our "connection graph", not just a static feed of people we know and things they post.
That's exactly what the Solid [1] (or as a sibling mentioned, SoLID - Social Linked Data) project is about. Solid proposes decoupling the presentation layer and the data layer, having data providers specialize in helping FOAF or other RDF data be stored and easily queried, while the view layer would contain logic that queries your FOAF/other-triple data and gives you the experience you are looking for in an app.
I was waiting for someone to mention Solid. Solid takes all the bad ideas from the Semantic Web and goes all in on stupid. They've been pushing WebID for years which is flawed at its very core. It ties a TLS certificate to a web resource, your WebID profile, but isn't signed, anyone with access to the server could steal your identity. When you mention that to them their usual response either, "No problem, a raspberry Pi is $35. I run one in my basement and you should too!" or "No problem. You can change providers if they ever steal your identity because with Solid you have choices!".
So you bite the bullet, order a Pi, register a domain, basically become your own sys admin - and you thought getting your mothers wifi working was bad - running your own little data center out of your basement. Now you need to actually use this WebID thing...it just happens to rely on the HTML keygen tag. Haven't heard of that one before? That's because it's a rarely used, on again, off again, supported tag. That's not really going to work so they decide to do what they always do, piggyback on something that actually works. Enter, WebID-OICD that piggybacks on OAuth2/OpenID connect except profiles still aren't signed, oops.
None of your information is encrypted so your pod hosting provider has complete access to everything. For some reason they're cool with that. There's no way pod providers would collude to aggregate your data. You're going to have so much choice and the tech is so awesome that you're just going to seamlessly migrate from provider to provider that it's never going to happen. Don't worry that doing anything with it is going to be hopelessly slow because everything is federated.
> Asking ever day users to create a site (buy a DNS and all that) and create a FOAF file and host it is kind of ridiculous.
You're describing a 2nd order effect created by behavioral dynamics all underpinned by the same core concept: ownership.
Your experience of usability, discovery, security, performance,... is entirely determined by the affordances of the tools you're using. And your level of control over those affordances is pretty much determined by the extent of your ownership over those tools.
If you use social media, you don't own the tools & infrastructure that host your data, including your social graph. You're basically a tenant on the lands of a liege lord and you pay by disclosing every aspect of your behavior to your lord.
Relying on cloud services is marginally better. Freedom comes at a cost. Leasing a VPS or shared hosting space, you're not bound to the strict terms & conditions of social media platforms. However, you're still a tenant to your hosting provider. Some people still don't like that. For one, the provider could go dark (go out of business, etc.) and you lose control over your data.
You could take it up a notch and lease a physical server in a data center. But that would still make you a tenant. Owning the physical hardware would get you even more freedom and control over your machine. But you don't control the data center. That's when you start to enter the realm of purists among the self-hosted community: self-hosting a physical server at home and leasing an IP address from your ISP.
Of course, every day users aren't interested in creating a website. But the vast majority of humanity didn't have a significant footprint online some 15 years ago either.
Billions suddenly started using the Internet/Web in earnest because social media made it extremely easy to connect and publish to wide audience at virtually no financial expense.
However, there's no free lunch. Never was. Someone pays the bills to keep the lights on. And those who pay, well, they hold the governance of those tools. And this includes how everything ranging from advertising to the algorithm that selects the types of information it's going to push in your eyeballs.
You have the right to pursue freedom, privacy, security, performance,... but you're not entitled to them. They come at a price. Either by doing the effort and actually doing everything yourself; or by paying someone to do it for you.
The promise of semantic web technology is that it preserves interoperability between data sources without having the condition that you need to host your data on a platform you do not own.
However, semantic web technology did not promise to solve the overall problem of ownership over hosting and publishing tools. This actually being a compounded problem which consists of the problem of intellectual property (open source), knowledge (learning how to code takes time and effort), scarcity of resources (hardware, electricity, physical security,... all costs moneys).
By and large, over the past 30 years a ton has improved. Unless you're a self-hosting purist, anyone can host their own website for 50$-100$ a year using popular VPS providers; reducing the problem largely to a matter of lack of technological knowledge on the part of the average user.
Then again, the vast majority of humanity uses social media not because there's an actual need to do so; but simply because it's there. Consider that asserting that people using social media equates "progress", may actually also just be hindsight bias. But that's another discussion.
> Asking ever day users to create a site (buy a DNS and all that) and create a FOAF file and host it is kind of ridiculous.
Plus it doesn’t do anything by itself. Someone still need to write software that will parse all those files, follow links, and present the information in a sensible manner. Somehow the SemWeb guys have this magical thoughts that once data are online in RDF the web will be revolutionized. Expected it doesn’t because writing software is the hard part, not converting data from a format to another.
Then there is triple stores. The other day I was in front of a Sparql endpoint, with no documentation so I couldn’t get any data from it. And if I got any I would still had to do heavy processing to make use of it.
SemWeb is just string programming with IRI to be fancy. Just let it die.
> Did Facebook simply get there first, or did they instead just do social networking better than everyone else?
Isn't that kind of obvious? The article kind of answers that itself:
> In the beginning, way back in 1996, it was SixDegrees. Last year, it was Friendster. Last week, it was Orkut. Next week, it could be Flickr.
Friendster, Orkut, Myspace, Google+ (ironically, I had to Google the latter, as I couldn't even remember its name). Many social networks have come and gone.
That's not to say Facebook hasn't likely done unsavory things to perpetuate its dominance. But it was better, in one or many ways, than all that came before it, and all that have tried to succeed it. Many haven't even tried. Twitter and Instagram don't even attempt to replicate your real-life connections. Hell, I can't even remember what pseudonym my real-life friends are using on these other networks. They're playing a different game.
This is a strength for those networks, in some ways. For those who care about what "influencers" say, or in what opinions are shouted the loudest by the folks followed by the most loud-shouters. If you care about both a footballer's opinion on a match as well as what he ate for breakfast; what a Kardashian wore as well as what she was paid to say.
In many ways, the various social networks aren't attempting to compete, they're attempting to 'win' their own niche.
It is very hard to gain traction now when Facebook already has a near monopoly for most but niches. That wasn't the case back when Facebook started.
I'd actually think you can say that Facebook simply got there first, and with first I mean to first capture the large swath of people that wasn't techies nor young people. They had great timing with when the mainstream started to get pretty comfortable and curious about the Internet (which was a pretty huge deal considering how big the barrier to computers and the Internet was for many/most people).
MySpace had a HUGE lead. Facebook came in with a superior product and won on its own merit. If you want to claim first person advantage, you should definitely point to Friendster though in the US or Orkut in South America.
However, for all the dozens of social networks trying to unseat the behemoth MySpace, they all collapsed under their own weight. Turns out building a scalable network of social connections is a very hard problem to solve.
I think parent's point still stands. They're saying it's hard to compete now
I don't think it was intentional on Facebook's part, but in retrospect it would have been a brilliant strategy. MySpace was largely populated by a high school/shortly-post-high-school crowd.
Facebook came along and your college had to be added/invited (since residence halls and class schedules were curated). Nothing drives demand like exclusivity. Now all of MySpace's former demographic was clamoring to get in -- either by graduating HS and moving to college, or by having their existing college added.
College-only Facebook would have died after everyone graduated and it really took off after they decided to compete with Twitter's feed. The model of having a stream of news of people complaining about what they had for breakfast allowed grandparents to get massive FOMO and join to see what their grand kids were up to.
Zuckerberg for all his flaws has always had an eye for what the next big social media wave is and went after it.
Agreed, but I think a(n unintended) side effect of college-only was to drive the demand.
Things we utterly hated, like turning a pull-only format of "let me view my friend's profile" (which later became "what's changed is in yellow") into the push style "newsfeed", and requiring real names, really made it what it is today. Obviously there are still fake accounts, and people aren't all using their real names anymore, but it's so much closer to a "real" profile than anything else out there.
But yeah, especially the newsfeed. Making users poll friends' profiles never would have resulted in the engagement they require. Frankly, the platform would have died because user interest/engagement would have gone elsewhere.
I used to agree that Facebook was superior, but I don't anymore. Myspace allowed people to far better express themselves with design and music. It was a shit show, but so are we.
> Facebook came in with a superior product and won on its own merit.
I don't know about superior but it didn't allow one to butcher their page with custom themes such as yellow text on white background. It became a cesspool of awful web design. Facebook was simple and clean.
Awful or not, the ability to customize themes etc encouraged a lot of folks to learn basic web dev. The level of 'control' goes a long way towards the general point the article is making. All that has been lost for the sake of a 'simple' and 'clean' interface.
I just googled a little bit (won't share the links because they are easy to find and check, and I am on mobile, at work).
Orkut had ~29 million users by the time it was passed by Facebook in number of users in Brazil. This was about ~65% of internet users in Brazil.
Today Facebook has ~75 million users in Brazil, which is about 53% of internet users in Brazil today.
Of course, it all depends on how you define "monopoly" here. But with the definitions: "% of users that use any social network" or "% of users that use the internet" , Orkut had as much a monopoly as Facebook today. If the definition is "% of users from the whole population" that probably Facebook has more today.
>That's not to say Facebook hasn't likely done unsavory things to perpetuate its dominance. But it was better, in one or many ways, than all that came before it, and all that have tried to succeed it.
They understood that multi sided markets can be "self promotional", people (or advertisers) get something out of it, therefore the actual users, advertisers and those who run communities drag more and more people in to get even more out of it.
Back in the mid-90s to the turn of the century I was fond of and used Bolt.com, which might be described as "Facebook 10 years ahead of its time". Users had a profile page with their own username, picture(s), and related information, as well as access to mid-90s era style forums to live-chat with other users. It died around or just after 2000.
On the subway tangent, Tokyo does continue to have multi-company competition in its transit system, yet has continued to build out its network, including new lines through the middle of the city in 2008! (Fukutoshin)
There’s a lot about why the system works, but it goes to show that network effects aren’t insurmountable
It’s a bit more costly then some alternative systems, but it’s better to have an expensive line that exists rather than a cheap line that doesn’t.
As does Singapore (MRT and ComfortDelGro compete), which is arguably one of the most effective and cheapest systems in the world. There isn't infrastructural problems, (as I understand it, I could be wrong) as a part of the operating contract, they are required to have interoperable fares.
The piece starts with an analogy about the MTA, but I find it a rather weak one. The NYC government could have imposed a regulation that any subway built in the city must be built by certain standards for tunnel size and the ability to interchange. That would have helped provide some ability for the two systems to interlink. Moreover, the city government intervened significantly into the operation of private train and subway businesses. The article even notes the city tried to compete directly. Government-run businesses often use the law to squash competitors. Then, when the competitors go out of business, they say it's a natural monopoly and everyone forgets about it.
> This issue of identity was an acute one for FOAF. (...) there does not exist and probably should not exist a "planet-wide system for identifying people," (...)
> Do we trust the homepages and conclude we have two different people? Or do we trust the email addresses and conclude we have a single person? Could I really write an application capable of resolving this conflict without involving (and inconveniencing) the user?
It seems the author identifies identity online as a major problem.
Interestingly, the last years have witnessed significant efforts towards a concept called "self-sovereign identity" [1, 2] that tries to solve exactly this. Emerging standards at the W3C [3, 4], open source software at the Hyperledger Foundation [5], and even non-profits [6] are working globally on the technology and governance of such a system.
Eventually, I think, self-sovereign identity will happen. I wonder what will be the impact on decentralized social networking.
Around 20 years ago, friends in college used this French social networking website Zdarmanet. We mostly used it as a distributed contact list. I often wonder what it could've been if they had been a bit more aggressively commercial.
Realistically though, the strong (at the time) French database and computer privacy laws likely would've prevented it from every doing the slow-boil takeover that Facebook accomplished. In the end, even those laws were ineffective in handling the foreign behemoth that Facebook became.
Edit: website is now closed, displaying "After more than 15 years I am stopping this service. Sorry... in case you would need anything related to that please see notojo.cz for how to contact me." and I guess it was of Czeck origin, not French?
Distributed things are more resilient, but centralized competitors tend to scale faster.
Once a given centralized competitor has scaled sufficiently they will achieve a monopoly due to network effects and predictably abuse that monopoly to reduce future competitors. Tale as old as time.
Friend of a Friend: Boy was the internet a different place in the pre-Cambridge Analytica era.
Seriously, I can't see how something that revolves around the assumption that everyone would just merrily broadcast their entire friends network to the whole internet would be able to fly nowadays.
Legislating natural monopolies way after they were obvious has been a problem for more than a hundred years. It is probably a case where the free market leads to a problematic situation and hence it is hard to discuss for ideological reasons.
Note that the social network Hi5 used (still uses? not sure) FOAF. So it can live harmoniously with centralized networks, if the networks choose to make use of it.
it’s amazing how much quiet your life becomes after you unfriend 99,9% of your social graph, bit similar to noise canceling on the airpods. give it a try :)
Quoting from Wikipedia:
> Since communications services exhibit network effects and positive externalities, new entrants would face barriers to entry if they could not interconnect their networks with those of the incumbent carriers. Thus, another key provision of the 1996 Act sets obligations for incumbent carriers and new entrants to interconnect their networks with one another, imposing additional requirements on the incumbents because they might desire to restrict competitive entry by denying such interconnection or by setting terms, conditions, and rates that could undermine the ability of the new entrants to compete.
So either social media companies should be classified as telecommunications companies (imho not far-fetched), or we need something similar to the Telecommunications Act but for social media.
[1] https://en.wikipedia.org/wiki/Telecommunications_Act_of_1996