Jack Dorsey Unveils Bluesky Social, the Decentralized Twitter
geekmetaverse.com
geekmetaverse.com
The second is that we seem to be seeing a lot of VC brain here. By that I mean, I can scream at a VC "OMG THERE'S A TRUCK COMING AT YOU AT 100MPH" and the VC will respond "Well what we need to do is look at how the concept of a truck is cha-" before becoming a fine pink mist. I'm not impressed by people who can't answer basic questions about their product. I know Jack will hate to actually address this, but: how do you balance speech, censorship and moderation? I get we can wax lyrical about freedom of speech, but how are you actually going to address thousands of people shouting the N word at your users? And if your answer is "It's decentralized, it's down to X,Y,X people" then I have to point out, you've just outsourced the most difficult differentiator of a social media startup.
The real question here is defining "public square" legally.
Mobile phones utilize IP networks (VoLTE or voice over data networks), so dialtone 2.0 becomes an easy option with domestic network protocols as well.
ESIM is about to change dynamics as well, so any use of a phone number/SMS/2FA for identity verification is rendered null with the current system in place.
https://consumer.ftc.gov/consumer-alerts/2019/10/sim-swap-sc...
Domestic networking protocols and dialtone 2.0 easily address most ransomware, malware, SPAM, etc much easier than trying to police 3rd world countries.
That severe limitation of regulating 3rd world countries is also the same argument with climate change. The can continue to regulate their own soverign borders.
Almost none of them in competent jurisdictions even attempt to define spam with an eye towards balancing against free speech.
Here's the Canadian jurisdiction against SPAM: https://laws-lois.justice.gc.ca/eng/acts/E-1.6/index.html
> An Act to promote the efficiency and adaptability of the Canadian economy by regulating certain activities that discourage reliance on electronic means of carrying out commercial activities, and to amend the Canadian Radio-television and Telecommunications Commission Act, the Competition Act, the Personal Information Protection and Electronic Documents Act and the Telecommunications Act (S.C. 2010, c. 23)
> Unsolicited electronic messages 6 (1) It is prohibited to send or cause or permit to be sent to an electronic address a commercial electronic message unless
> (a) the person to whom the message is sent has consented to receiving it, whether the consent is express or implied; and
> (b) the message complies with subsection (2).
And here's the European jurisdiction against SPAM: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A...
> Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector (Directive on privacy and electronic communications)
> Unsolicited communications 1. The use of automated calling systems without human intervention (automatic calling machines), facsimile machines (fax) or electronic mail for the purposes of direct marketing may only be allowed in respect of subscribers who have given their prior consent.
Imagine there have a platform run as you suggest and I don't like what you're saying on it: I flood you and anyone that supports you with not-quite-illegal abuse, misinformation, non-sense, spam, false flag support from people with deplorable views. Copy-cats of you that take your positions but weakly argued or with a touch of added genocide. At the end of the day almost no one will see what you wrote, and most of those that do will think you're probably a kook with deplorable views (since that's what you're surrounded with).
Censorship that works by outright blocking speech is a largely outmoded concept in a world where communications networks connect every one to everyone else. In such a world flooding with noise that undermines the conversations people want to have and undermines their willingness to attempt the conversations is a more effective tool to censor communications. Outright blocking isn't very effective, unless it's done secretly, because people will route around it and plugging leaks is extremely hard (requiring essentially AI complete filtering).
As such, censorship is now essentially its own opposite.
This means there really are no magical silver bullets. We ought to stop pretending that private spaces are public squares though-- though this is a problem offline too: with eminent domain being used to seize land and create giant privately owned open air shopping malls in many places... but at least offline doesn't have the same scaling properties that make it so easy to censor by inundating with noise.
But I honestly don't understand this whole argument. Twitter is a social network, you just need to only receive messages from your network or "friends of a friend".
You can't have both a global network and get offended given the multitude of opinions and laws in the world.
The problem seems to be less that people are offended by something, but that some people don't want other people to even be exposed to some thoughts. Some paternalistic higher-than-though attitude.
Even taking at face value what you just said, it doesn't logically follow that if having one or more places like 4chan is just fine, having online venues be like 4chan would be just fine as well.
4chan is actually heavily moderated. It may just not be legible to you because they're not removing the stuff you're expecting but different stuff instead.
There are factors like topicality, being small sized, or being highly ephemeral which make moderation easier or lower amounts more viable-- but twitter is pretty weak in those respects, sadly-- I think that's how it manages to both have a reputation for being toxic and overly moderated.
Working fine at doing what? Emulating a thousand typing monkeys? Arguing that 4chan has some form of working discourse and rherefore everything should be like it is like arguing there is still some message at the end in the childrens game "silent mail" and therefore it should replace all methods of communication.
Don't get me wrong, there probably even is some sort of use in the existence of something like 4chan (I bet some sociology doctorate has written about the function of the platform for certain demographic parts of the population for example).
But there is probably a reason why you are discussion this topic here on HN instead of over there. And if I guess, this reason has all to do with you being able to have better, cleverer and more informed discussion here, than over there and is not at all hindered by the fact that the HN moderation would cut your voice off if you just pressed the right (wrong) buttons.
Some form of moderation is always better for discussion than none at all, unless you are making some sort of social experiment.
It's already hard enough to keep up with the laws in software.
Sort of? Centralization is more profitable and easier than building new open internet protocols and formats.
Computer code still runs on CPUs and not the aether. If its accepting connections from other computers it is the server in the client/server model. If it's on your desktop, it just typically means it doesn't have ECC memory and a battery backup dropping the reliability even further. Decentralized architectures have all kinds of fun and complicated failure modes, and especially for consumer use, high speed and low latency typically are not their strong points.
Usenet, IRC, email - discussion, conversation, mail - all worked fine for decades on the old, low speed internet. What exactly are we getting in exchange for all this high-availability nonsense anyways?
So yea, we're the old men yelling at clouds, and the kids have gone on to a ADHD fueled media filled flashing light fest and they don't plan on coming back.
The good part is that operating a usenet server with no alt.binaries.* content is extremely low demand; it was designed to run on the 70s internet, so...
Likewise Usenet - yes, we abused it to get our alt.binaries (read: boobies), but that was strictly desperation. It still works fine on the fast internet, but again, will continue to work on a 90s internet, in the event that we lose the fast one.
Fast itself is really what's warping our perceptions of what is normal and acceptable. When was the last time you watched an old Connery Bond movie? My grandpa took me to see every Bond movie that existed at the time back in the 80s, the downtown cineplex threw a festival. I remember enjoying all of them at the time, nothing wrong with em.
I threw on Dr No a couple months ago and I swear, I could not get past about 30 minutes, the pacing was hellishly slow, to the point that wanted to climb into the screen and lob a bomb at his car myself out of sheer boredom. Had a similar experience trying to watch Three Days of the Condor.
Interestingly, I decided to rewatch the Lone Wolf and Cub movies after watching the Mandalorian, and they were just as fun as I remember. Some slow scenes, but good pacing with a decent body count:frame ratio.
I don't think we're so much yelling at clouds as wishing the clouds would slow down long enough for us to yell at em. I have my doubts that the pace can be maintained indefinitely, and they are not trying to bring it under control at all. What I expect next is a painful correction in which the laws of nature will come into play in a big way. Whee!
But say you track 1000 peers, find 50 online, and send a 280 character tweet to ten of them. Sure it's going to take seconds for that tweet to get to whatever fraction of millions of people that are interested. Sure your client should ask the 40 peers if they saw the message to make sure it went out.
The beautiful thing is that even on a smart phone paying 10x the bandwidth of a tweet is still effectively zero.
If the goal was a commodities trading platform, sure distributed wouldn't make sense. But receiving tweets of interest send in the last few minutes seems rather insensitive to a peers bandwidth, latency, CPU, or ram.
Decentralization does help a lot here. The X, Y and Z server owners don’t have to agree or coordinate. The fact that they will disagree on their policies is part of the answer. There won’t be a one-size-fits-all policy. Different servers will try different things and ideas will be shared about what works and what doesn’t.
Z server can choose to allow saying mean things and X and Y server can block and ignore posts from Z.
Some servers may have a political bias one way or the other and individuals can choose which to join and pay attention to.
Back in the Usenet days we had killfiles to do user based filtering -- everything would get sent out, and users would decide what sort of content they didn't want to see. Obviously it was harder back then to just spin up a new identity, so it's not simply a matter of taking the old tech and slapping it on the new, but the approach avoided the splinter effect we get trying to decide for users what will and won't be available to them (whether they like it or not). Snowden actually tweeted a week or two about how trivial this would be to implement on Twitter if they ever actually wanted to (with a working one liner to prove it). The issue thus far has usually been that advertising funded social media THRIVES on the controversy, and only moderates what it needs to to keep advertisers happy. They don't WANT you to make it a pleasant experience, because controversy and fights get more ad views.
Perhaps at some point bandwidth becomes cheap enough that we can cut back on how much of a role advertisers play in steering the ship. Until then I don't see a great solution to any of this.
Many crypto scam startups have made bank with this less efficient strategy. Have you seen all the companies selling banks on internal proof of work blockchains and stuff?
Twitter has two killer features: the people who use it, and search functionality that allows you to get near-instant results. Mastodon (and I assume Bluesky) cannot compete on either.
Privately accessible content tends to be much smaller so local caching works. And if communication is not meant to be instant, email proves that a decentralized system works.
What gets a little more difficult is content that is widely available but behind login. Facebook has this but twitter and mastodon tend to not be this way.
In fact mastodon is a workable system for decentralized social networking, so I think the skepticism is unwarranted.
The only question in my mind is whether a popular social network can be effectively bootstrapped without a core capital incentive.
Or the difference between 95% of the results of a search and 100%, especially if the 95% is missing the least popular 5%.
“Twitter ex CEO announces half-baked and social signalling fuelled, competitor product to entice as many as possible to leave Twitter and help line his pockets”
And his ego. The world already has both Mastodon, a decentralized twitter-like social media, and ActivityPub, a protocol for decentralized social media services to connect to each other. Dorsey has chosen to support neither, and instead make his own service and protocol that can't connect to it.
Bluesky has a wait list, while there's already a google doc full of Blue Checks that have moved to Mastodon[1]. I'm hoping grass roots open source succeeds over a billionaire with VC funding, but we shall see.
My grandparents also talk about how wonderful the past was, but that doesn't mean a lick of it was true. Was MySpace better? In some ways yes, but that's because we were better as social media users, and such massive bot farms, and financial interests in corrupting said social media didn't exist.
If you turn on MySpace today it would immediately become foul shithole. The first thing some spammer would do is start creating profiles based on every identity they could find and the site would have to deal with that. Massive amounts of time would be used in getting rid of scams and spam on the sites. User content would not be neutral because peoples internet identities are no longer neutral, we've weaponized and politicized them. In order to maintain the neutrality you'd spend untold amounts of time banning content and polarizing a large amount of the user base against you turning them adversarial.
Social media is a garbage fire these days. Most of the time I try to avoid talking about it on HN in order to stay in this sites guidelines.
The issues we have with todays social media are the same issues IRC, usenet and other primordial internet social tools have suffered from in the past at scale. People abandoned those for what they thought was better. They went on to myspace and eventually they gave up myspace and went on to facebook and twitter because they thought these would be even better. And they were for a while. Facebook was really cool when people limited to sharing meal and travel photos with friends and family. Twitter was cool when limited to sharing fun short stories or updates.
Then gradually the noisy plebe invaded. The drunken pub types debating angry politics, pseudo intellectuals, wannabe thought leaders, cheap scammers and sleazy politicians.
And instead of banning or pushing them towards network ghettos, algorithms where made to amplify specifically their voice, in order to bait all the decent people into engaging with platforms even more. These algorithms developed habits in us that turned everyone into a little bit of the above.
We tried this. Doesn't work. We need to iterate what worked. MySpace did. I fail to understand why facebook and twitter didnt put more thought into how to monetise the platform in better ways. Was it lack of imagination or the lack of people skills of their founders? Anger and politics are not the only things that drive sales and profits. Social networks are a reflection of our society and in a society you tend to marginalise the angry violent mobs that try to capture your attention and suck your energy, and instead you go where you can do your shopping, eating, reading, traveling, entertainment consuming safely and peacefully. Social networks can monetise on that.
Instead they made us fight each other.
Like you say, myspace was actually quite good, but users
>gave up myspace and went on to facebook and twitter because they thought these would be even better.
People may be questioning this now more than ever, but it's too late for myspace now, and only hindsight will pinpoint peak Facebook and peak Twitter after those milestones are unmistakeable.
These are not just "social" networks (some say anti-social), they are pop-culture trends.
Which respond to and influence culture as much as they can while they are popular, the more popular the more so while it lasts.
Capture the beauty of life not the garbage of it. Look at human psychology and see what else other than anger can increase engagement and what algorithms can create a better culture. There is more than that emotion that can drive growth. I refuse to believe we are all angry shits, we can do better.
I'm very into this stuff. I was into scuttlebutt long before bluesky was announced even.
But the problem I see is that with ActivityPub I already have a whole ecosystem of software and libraries, while with ssbc I have no idea where to even start. When ssbc was first announced I tried using it briefly but it was very basic, I hope they've come further.
With the fediverse I can literally jump straight into a container and have an instance running within minutes.
Is there an equivalent to activitypub.rocks for sscb?
https://ssbc.github.io/scuttlebutt-protocol-guide/ maybe this guide.
Bluesky is not related to secure scuttlebutt, they are working on a new protocol currently called "AT Protocol". https://atproto.com
ATP is mainly federated so a bit closer to mastodon/matrix than SSB which is quite a bit more decentralized iirc. Do check out their docs, it's a work in progress but should give you a better insight to what they're doing.
1) you can move your identity from the gitgo. Yeah, sure, AP allegedly allows this, but your are at the mercy of the server owners to cooperate. Assuming, of course, the server even is still reachable. And even then, recent (last 5 months) there are numerous examples of transfers failing badly.
2) This protocol doesn't force you to trust that the server admins aren't reading your messages. This is not the case with AP.
3) The relay system, whereby to get notices from other communities, is wildly broken, and the two largest Mastodon servers don't even support them.
I'm as skeptical of Mr. Dorsey as anyone, and the community absolutely deserves the right to see deep into this protocol. That said, implying that Activity Pub meets or beats everything this new protocol offers is not correct.
I was hoping since identity is such a corner stone of this project that there would be a more fleshed out auth system (OIDC?).
Other than that it seems like a reasonable implementation, but I don’t like the addressing scheme.
They are using identifiers that look like emails: bluesky@bluesky.app
But they are not actually emails so they can collide with real emails.
The auth schemes will evolve. That technique is meant for your main app.
I guess you've been ignoring watching Google squirm for years about moderation, they left China because of this. They've been fined for years in the EU for violating laws around this and privacy and this is only expected to get worse and more expensive for them. Moreso Google has deleted plenty of individuals/entities from the Googlesphere for questionable reasons that's leading to increased resentment over the years. Google hangouts/+ epically failed.
Google hasn't pulled off anything in the social media market other than losing money
There was the whole creepy men in diapers videos ending up on kids YT for months on end (Elsagate?), and mass demonetization issues if the topic is even close to controversial (of course they still show ads to you, the creator doesn't make income). Along with bans across all types of topics that are leaving more and more people confused. When you add in recent pushes for YT shorts because sites like Tik-Tok are pulling lots of viewers away, I still don't think the picture is near as rosy as you make it.
The one thing that YT has is a near monopoly on video serving and a 'special' relationship with the movie and music industry that prevents them from being sued off the face of the planet.
I would argue those are 2 of the most pressing issues social media faces. Maybe others would disagree however.
I would also say that phones are a bad example because while the phone companies are decentralized, the call routing systems have written law behind them to ensure ATT and Verizon actually route calls to each other (and all other companies like CLEC/ILECs).
Your endpoint might be popular and heavy used or you could be self hosting and be the only person using your endpoint. It's decentralized and works for the average person
Mastodon is part of a federated network. Most people who use Mastodon belong to a large server rather than hosting your own, but you can be part of that network without being part of mastodon.xyz. You can join mastodon.social, or any number of those mastodon servers, or host your own. The same is currently true for email.
What does it mean to be blockchain optional?
I notice that the documentation[0] says:
"We introduced DID Placeholder because we weren't totally satisfied with any of the existing DID methods. ... We cheekily titled the method "Placeholder", because we don't want it to stick around. We're actively hoping to replace it with something less centralized. We expect a method to emerge that fits the bill within the next few years, likely a permissioned DID consortium.
which seems to be an admission that the hard part of distributed social networking isn't actually solved yet, and the project relies on "actively hoping" that a solution will eventually "emerge". I'm also not sure what a "permissioned DID consortium" even is, let alone how it would be useful here.
Anyway, it seems a bit premature to abandon ActivityPub and introduce yet another incompatible social networking silo, based on the hope that the new project's magic placeholder beans eventually grow into a sturdy beanstalk. Perhaps "Beanstalk Social" would therefore be a better name for Jack's project (and he could even keep the same acronym).
The DID PLC registry is a placeholder because what we actually want is akin to a ICAN-style governance over it. When we talk about a “permissioned consortium” we mean as opposed to a “permissionless blockchain.” It’s not that we haven’t solved the mechanics; it’s that the right solution is a people/org problem and not a computing problem.
We include DID Web as an option in our system, which comes with its own tradeoffs but essentially falls back to DNS+HTTP as a replacement solution. We did that for a lot of reasons, but one in particular is understanding that we may indeed have gotten this technical decision wrong, and need to move away from PLC. I should note however that DNS is akin to what we’re aiming for with DID PLC — a global org operating the central registry.
The most egregiously absent solution with PLC is economic: it costs money to host the DIDs, but we don’t want to pass that cost on to end users. We’re going to need to suss that out, but the answer might just need to be that server operators pay for their registrations.
I can understand you preferring ActivityPub. We wanted to pursue a different design. If we’re wrong, and AP thrives, I’ll be happy enough.
Also, I missed “cheekily” in the copy review and I’m taking that out.
I think I am still uneasy, though, about the idea of hoping for a global governance structure (with unknown accountability and financial controls) to come along and start granting (and revoking) people their online identities.
I just feel like the professor in the Far Side cartoon pointing to the blackboard and saying "I think you should be more explicit here in step two"[0], but miracles happen every day, and important breakthroughs can take years of hard work before the answer is clear. Let's hope that you don't get stuck with a system that you're not happy with.
[0] https://qph.cf2.quoracdn.net/main-qimg-a3b19e99fad0ee2c1d663...
Very fair concern. We’ll keep communicating what’s going on there and see what happens.
Thanks, but no thanks.
So if I'm a social media platform that implements this protocol, am I restricted to supporting only the activities that Bluesky supports? Like if I wanted to have a network that distinguishes people I follow from people who I personally know, am I just out of luck because protocol doesn't support that distinction? And if I do create that distinction, what is its meaning when a user decides to migrate that data elsewhere?
The two situations are actually very similar, even down to the crowd that took over freenode when everyone sane left for libera.
stay tuned for the secret south african prince NFTs :)
Don't need anymore drama from these reckless people.
How will it be the same or worse? Which member of it?
Mastodon is simply a social federation enabling software; it's similarly abstract to refer to Twitter's software infrastructure.
Why is lack of moderation a problem on Mastodon?
Twitter pushes content from random users you don't follow, that creates a responsibility to moderate (plus they need to keep the advertisers happy).
The Mastodon timeline is direct follows. If I see some stuff I don't like, I can talk with that person or simply unfollow them.
and the TweakNewTwitter browser add-on rejects it, along with "trending" and all the rest. Add in ublock origin and you've pretty much got Twitter the way I thought it should be.
Having a process that distributes the load of moderation is better, IMNHO, than piling all that work on underpaid people.
The left 10% will hang out on Mastodon or Fediverse, the right 10% will go on Gab or whatever. I think what's happenning is great for all. You do you™.
Discussing Mastodon is hardly "forced" when the overall topic is a new decentralized alternative to Twitter, since Mastodon is a decentralized alternative to Twitter that's been around for a while.
It seems unlikely that this is astroturf given that the person who posted it has been on HN for over a decade (though admittedly not super-active) and their comments are on a wide range of topics and not just Twitter or Mastodon.
Let's see how that goes a year from now.
I was just stating that if twitter was bad before, these new alternatives will be 10x worse since they will create truly silos and radicalize even more people. To each their own.