Decentralized Social Networks Won't Work

For a federated service, like email or phone, this is not a problem: the union of all service providers is the entire service, and more providers can connect. So connecting to any compliant provider is fine, and providers have an incentive to be good, or at least outcompete each other by some metric.

For a siloed and sealed-up service, like Facebook or Twitter, once you and (most of) your social graph is there, there's no way anyone can connect from the outside; to be connected with you, they need to dive into the same silo. The bigger the service grows, the harder it becomes for people outside not to consider joining, because of the pull of the part of their social graph who already joined.

I think that federated services will always exist, as long as unimpeded internet connectivity is allowed. I also think that 1-2 huge walled gardens will also exist, for the same reasons why phishing using dancing kittens videos will continue to exist: many humans are emotion-driven and don't value [insert a list here] when overwhelmed by a positive emotion from something new, cute, and free to use.

Just look at Hacker News for example. It'll never be as big as Facebook, Twitter or Reddit, but it doesn't need to be. It's a good community for the people it's aimed at.

The idea you should only make stuff for as large an audience as possible or to get rich quick is perhaps one of the most depressing aspects of the internet (and world) today.

1. let sponsors host your data in an open competitive marketplace.

or

an open marketplace for goods and services that focuses exclusively on driving value to customers through:

1. require low cost guarantee from vendors.

2. vendor can't buy rank at all.

3. better way to find products and service.

(sort of like ebay that focuses on customer value not transaction volume)

do what amazon did to "best buy". you do alot of your research everywhere else but you buy from here.

Difficulty: the answers to the above must leave you with a product easy to use and reliable enough to compete with doing approximately zero work to use and maintain a Facebook account. Plus it needs to offer at least as many features.

I've for some time wanted a kind of private swarm-storage product (probably using IPFS, I guess, for lack of alternatives) that lets you plug in a new node at a friend/relative's house with a few TB of storage, put in the credentials (keys, bootstrap node[s]) for the swarm, and has it intelligently back up some amount of content from the network such that everything has at least X copies (3, say) while making all the rest available (that's where IPFS helps) on request, with some interface for adding content to the swarm from any node. The above problems have kept me from attempting even this task. A social network would be even harder.

A world where it's common for people to run their own server opens up exciting possibilities (just as a world where everyone carries a computer in their pocket has), but I don't think actual physical servers is going to get us there.

And yet, they are not limited to communicating with other comcast customers.

Whether that would be the best model for social networks, I don't know, but I think it is important to recognize that you do not need every individual to build their "server" from scratch to achieve some meaningful benefits of decentralization. If comcast doesn't like you, they cannot cut you off from the internet, and that is already a pretty big advantage over walled gardens.

So replace router with Apple TV.

Pointing out that some people run servers isn't much of a distinction. Some people also use decentralized social networks. But virtually nobody.

> It was merely not top-of-mind for the group I set it up for.

Some people make a big deal out of the "network effect" but they totally ignore how networks get going in the first place. It's the digital world, networks come and go all the time, overcoming the network effect is not a magical process. You offer more features than competitors and target people that have greater influence in the network.

It's for sure an uphill battle and this is one of the motivations of creating decentralised networks in the first place - to reduce network effects, to avoid these barriers to entry and these influences that make society's inequalities worse. But obviously if one has a depressing attitude about the whole process of overcoming network effects, and talk shit about "it can't work", then it becomes a self-fulfilling prophecy.

There's no shilling going on here. Just a user recommending it.

I do not believe a blockchain will solve it. The core problem is that you need a network of trust. Currently, we have governments as trust masternodes via identity cards. Then there are secondary master nodes companies like Facebook, Google, Telekom, Vodafone, etc, where verification is easier and faster, but less reliable. How does a blockchain improve anything about it?

Here is the challenge: There is a kid born in rural Africa. How could you proof to someone with a smartphone in the US that this human being exists?

Account creation costs real money, which we subsidize. Anti-spam is required. No phone number or email address is required to use the blockchain, only for Steemit Inc to buy you an account and give it to you for free.

Anyone with an existing account can create more, if they pay. AnonSteem simply sells creation for btc and ltc with a markup.

Whatever this might be, it's not theft. And it might just be something we have to accept if we don't want to live in a de-facto police state.

> Conversely, a lot of valuable content comes actually from reposting and "karma theft" (which is what linking e.g. a blog article from someone else actually is).

That's a fallacy, though pretty common. People who connect other people with stuff those other people want do provide a useful service, and being rewarded for a useful service is not obviously a bad idea, let alone some sort of "theft".

> Fourthly, how do you avoid fraud as for example often found on Kickstarter? Kickstarter actually has a review process and even they are not fully equipped for it. I am not seeing reliable moderation ever scale enough such that it can withstand the likely immense fraudulent interests due to the monetary incentives.

Just like everywhere else in life? You don't need a big brother managing every step you take in order to avoid fraud.

I did however find out that the community is great, made many new friends and read a lot of interesting posts. Lately it looks like it has been kind of overrun with people from China or third-world countries, which is interesting, but sometimes the posts are not translated and the steemit.com site does a poor job of managing different languages.

I haven't been active on Steemit for more than a year but thinking of going back since I miss it. It's very consuming though, a lot more than Facebook. Since there's a monetary reward on everything you do it feels like "working" even when you're mindlessly upvoting stuff. I used to spend many hours a day just reading, jotting on a new blog post or connect with fellow steemians on the official chat server.

You're going to have to do better than draw some analogies to monarchies on Hacker News. And things like "but they sell your data!" has had no impact beyond desensitizing everyone with Chicken Little'ism.

A good social network shouldn't be about who has the most users, but who has the users you want to interact with.

Mastodon will probably never be the average user social network, but it works pretty well for people with specific interest, or marginalized people that doesn't want to interact with hateful people.

As for the lolicon content, it just doesn't exists except on specific instance, which are blocked on mine.

If you think the only reason to use a self-hosted social network is if you want to circumvent the laws, you're making the same argument as only wanting privacy if you've got something to hide. YouTube ran an algorithm to delete "extremist videos" and accidentally irreversibly deleted a bunch of historical evidence of destruction of Syrian historical sites. Shit like that happens all the time when someone else is in charge.

I don't know about your definitions, but in my playbook 800,000 users is pretty damned successful. Let me reinterpret your comment for you: "I do not want to believe Mastodon is successful, since it hurts my view."

> Apparently content that was not accepted on twitter, such as lolicon, has found a new home on Mastodon.

Unsurprising. All communities end up with some version of this problem, be it hate speech, illegal content, piracy, etc. And it can definitely be harder to police in these kinds of federated environments... but it's certainly not an unsolvable problem. People just have to want to solve it. Twitter has a huge problem with hate speech on their platform, as an example, but they've basically turned a complete blind eye to it for years and years. (And even now after they've started trying to add some support for content policing, reporting hate speech still requires far too many clicks to be reasonable, given the rate people can churn it out...)

In fact, when your platform starts to have these problems, it's usually a sign you've done something right. It means your platform is actually mature enough to abuse, unlike other platforms that failed to reach this status like Diaspora. That's not a sign to give up, it's a sign that you need to start thinking about how to solve these new problems. (And they're hard problems to solve, often coming down to jurisdictional boundaries and "default deny" policies simply because the problem is seen as too hard to fix in software. You even see Google struggling to solve these problems - e.g. on YouTube it's still common that they ban content erroneously, claiming it to be extremist or whatnot.)

> The average user would never bother to migrate to decentralized services for any other reason then necessity, lack of choice, or to escape rules/laws.

Or you know, because they want to. There are plenty of reasons to leave Twitter behind: It's adding bloat that hardly any of the users care about, with yawn inducing, bandwidth destroying live video, infuriating autoplaying video, and "moments". It continues to pollute feeds with garbage from other users with no way to filter it (like "[so and so liked this]" - good for them, but this isn't Facebook...). And now they're even messing with chronology, so you get messages from hours in the past appearing at the top of your feed instead of up-to-the-minute content that you'd expect. (And my personal favorite anti-feature, the "you have location tracking disabled, we will not track your location" location-tracking sidebar filled with location-specific content.)

Looking at all of those downsides, then comparing with Mastodon which doesn't have any of those problems, supporting large messages, spoiler tags/content-under-the-fold, having decent Android clients, and decent, growing communities, and, well... it's really quite a compelling platform for me.

Imagine if we didn't have a standard protocol for emails (SMTP/IMAP/POP) but instead several protocols built using JSON on top of HTTP. We would have to take most of our time to work on broken transport libraries and fix incompatibility issues between all those solutions. Hopefully we have a standard (even if it's not the best one) and we can "just" build on top using the many existing libraries and projects out there.

I think that such "social standard" should be build on Internet protocols (TCP/TLS/DNS/IP…) and not Web only (HTTP), even if it's "easier". Also instead of re-starting from scratch I prefer to improve something existing like XMPP that have flaws but is already an industry standard with a strong community, many libraries/clients and servers deployed all across the world.

It is equivalent to a text message or phone call.

Those are probably locked accounts.