Dave Troy on Bluesky: A Rebuttal.

[EDIT, November 4: Improved formatting and corrected for spelling and grammar. Also edited language slightly for clarity.]

[EDIT, November 10: Improved formatting and corrected some typographic errors.]

Recently, Dave Troy posted a Medium article about Elon Musk, Jack Dorsey, and Bluesky, claiming a sinister connection between them. This article has gotten significant traction on Twitter, but has had very little in the way of critical scrutiny.


Let’s be clear: I do not like Musk. He is a narcissistic sociopath with an impulse control problem, who steals credit and ideas from others and seems to be perennially incapable of a truly original thought or joke.

On the other hand, while I like many aspects of (pre-Musk) Twitter as a company, the site itself isn’t one of those things I have ever liked:

In short: Musk is a cesspit of a human, Twitter is a cesspit of a site, and they are made to destroy each other. As of this writing (November 2, 2022), Musk is all but certain to run it into the ground, having fired both the top executive and the board of directors and greatly curtailed moderation, and currently planning to fund the site through turning user verification status into a paid feature any account can buy.

Dorsey’s continued support for Musk is puzzling. Perhaps Musk has convinced Dorsey that he is indeed a good fit. Then again, Dorsey is a crypto proponent, so maybe he’s just very vulnerable to this nonsense.

I thus have no love for Musk or Twitter, and not much respect for Dorsey. But Mr. Troy’s article, despite criticizing all three, is just … wildly wrong. It presents people like Musk and Putin as impossibly cunning, prescient men whose plots somehow defy the unpredictability of the future, rather than as the impulsive, sociopathic assholes they are. It elevates Dorsey to the role of a knowing accomplice with little to no evidence to support it. Worst of all, it grossly misrepresents both Bluesky in particular, and decentralized, federated, or open source social media (which I will call DFOS for short) in general.

I’m pissed off enough, so I decided to write this and post it. Let’s see what happens.

Part 1: The biggest error Mr. Troy makes.

The article is formatted as a weird cross between a frequently asked questions section on a web page and a Socratic dialogue. It opens as follows:

Q: Jack Dorsey is launching “Bluesky,” a new social network to compete with Twitter. I’ll just join that instead!

A: Sorry to disappoint, but Dorsey played a key role in Musk’s deal to take Twitter private. The two are good friends. And Bluesky is an initiative launched by Twitter. In April, Dorsey wrote, “In principle, I don’t believe anyone should own or run Twitter. It wants to be a public good at a protocol level, not a company. Solving for the problem of it being a company however, Elon is the singular solution I trust. I trust his mission to extend the light of consciousness.” [Update: Dorsey did not transact his shares in Twitter, and now holds equity in Musk’s venture.]

This first question-response implies that Dorsey himself is launching the platform. This implication is wrong. The CEO of Bluesky is Jay Graber, and Dorsey does not own Bluesky. Dorsey is merely on the board of directors, along with people like the inventor of XMPP (which the basis for a lot of instant messaging protocols). This is public knowledge and has been since it was announced. Mr. Troy should know better than to make simple factual mistakes like this.

The fact that Dorsey isn’t the owner or CEO of Bluesky, and does not control it, already undermines a lot of what Mr. Troy asserts in the article. However, this article has many, many other issues which I need to address; it is best that I go through this thing thoroughly, because it gets a lot worse from here.

Q: Uh, if “no one should own or run Twitter,” why did Dorsey advocate selling it to Musk… I’m confused?

A: To get it out of the hands of Wall Street investors, and turn it into a “public good at a protocol level, not a company.” Dorsey and Musk believe it can do more good for humanity if it’s an open technology than if it’s a company owned by any one person or by Wall Street investors trying to maximize profits for shareholders.

Now, Mr. Troy has not established (and in fact never does establish) that Musk shares Dorsey’s ideology on this. However, the rest of the answer isn’t that inaccurate; indeed, the answer sounds quite positive as descriptions go. What’s the problem with open technology, after all? Many people (myself included) think DFOS is a good idea, for a variety of reasons which I discuss later on.

Returning to topic: as said, there’s nothing sinister about open technology or protocols, and while over the years, some finer issues have been raised, it’s hard to honestly look at either of these ideas as being bad, let alone sinister.

Which is why Mr. Troy, knowingly or not, proceeds to be intellectually dishonest.

Part 2: Misrepresenting DFOS

Q: What do you mean, a “public good at a protocol level?” What even is that?

A: The foundation of the Internet is built on protocols that we mostly take for granted now. TCP/IP and HTTPS enable the web. Modern e-mail is built on top of SMTP and IMAP protocols. Usenet newsgroups are built on top of NNTP. Dorsey envisions a foundational “Twitter” protocol that anyone can implement and run. This would need to become a formal Internet standard, which requires going through the Internet Engineering Task Force’s “Request for Comment” (RFC) process. This can take a long time. This is what Bluesky has been working on starting to map out.

Here Mr. Troy shows that while he does seem to understand (or at least parrot) information about Internet protocols, he’s somehow not fully aware of the difference between the Internet and the Web. In particular, it is almost always the case that social media, regardless of form, is implemented via the Web rather than a dedicated Internet protocol, since social media pretty universally involves hypertext, videos, media, and other things that by definition are generally handled in Web browsers and the like.

Q: This all sounds crazy. Doesn’t Musk want to make money by owning Twitter?

A: The goals are more ideological in nature. Musk and his backers believe that the global geopolitical arena was being warped by too much “woke” ideology and censorship, and wanted to fix that by first restoring voices that had previously been silenced —and then implementing technical and algorithmic solutions that allow each user to get the experience they want. They think this can “solve” the problems that people cite about social media content. Making money, they figure, will come from the secondary effects of enabling “free speech” and the possibility of building other services like payments and replacing government on top of such an app. Plus the company’s social graph data is a goldmine for other businesses that may wish to benefit from detailed knowledge of the makeup of society.

Q: Uh, isn’t it naive to think that there’s a technical solution to harmful content? Isn’t that likely to lead to more radicalization and cultish behavior?

A: Yes, it’s completely naive and they are being willfully ignorant of the harmful side effects of content. It is a kind of tech fundamentalist solutionism that posits that for any difficult problem, there must be a technical solution. Many sociologists and cultural scholars would argue differently.

This part is … bad. Indeed, it’s so bad I barely know which end to pick it up from.

For one thing, it’s full of stuff that needs citation, particularly with respect to the claims regarding the motives behind Blue Sky. Remember that, so far, all Mr. Troy has to justify his claim about Musk and Dorsey being in complete ideological agreement is a series of leaked text messages, in which Musk’s responses seem to be his usual parroting. Let’s also point out that Dorsey isn’t the CEO or owner of Bluesky, but a board member. Thus, the phrase “Musk and his backers” asserts an ideological cohesion between Musk and Dorsey that might well not even exist.

However, that’s nothing compared to the mischaracterization of decentralized, federated, and open-source social media protocols that is displayed here.

For starters, there’s the thinly-implied assertion that these approaches are dangerous and haven’t been tried before. They have, and have even been wildly successful. Indeed, prior to the rise of modern social media, decentralized systems for internet chat were the norm, arising naturally from the decentralized nature of the Internet itself: Internet Relay Chat (IRC) was at one time the most common form of chat and still used by many to this day. It is inherently decentralized, with there being no one point of control for the protocol. Due to its age and certain other issues (IRC cannot natively display images or media) most people have moved either to Discord or to newer decentralized chat protocols such as Matrix; however, the resilience of decentralized protocols to takeover attempts was made very clear when, after a hostile takeover of Freenode, then the largest IRC network, in 2021, most of the traffic transitioned over to a new network, Libera Chat, run by the former staff of the old network who had resigned en masse.

Turning to more modern social media, Bluesky is not only not the first decentralized, federated, open-source social media protocol to be proposed, but it has established competition in the space that has been actively running for many years! Indeed, I am going to be cross-posting this on a Mastodon instance, which — you guessed it! — runs on ActivityPub, a decentralized, federated protocol involving a of other services and platforms, with 5.7 million accounts and climbing as of this writing. That’s right! Bluesky may ultimately have additional innovations as well as the backing of established talent and money, but it’s neither the first nor the largest of its kind, and in fact it has a lot of catching up to do if and when it does launch.

Even on closed-source, centralized platforms, there have been successful uses of systems where moderation is mostly if not entirely done by people other than a central team, usually individual users. Two particularly prominent examples are Discord and Reddit. Both of these services are divided into subsections (servers in the case of Discord, subreddits in the case of Reddit) and the overwhelming majority of moderation is done by the people who control these subsections. This allows for each subsection of the site to set its own rules in addition to the site guidelines and enforce them accordingly. As a result, rather than having to rely on a centralized team of humans, or on AI moderation with its high false positive rate, individual users, in communities they control, can rely on themselves and on each other.

Secondly, Mr. Troy is excessively dismissive of … well, a lot of things. He dismisses the idea of decentralized, federated, open-source social media as being “tech fundamentalist”, claiming it tries to solve the problems of moderation by a “technical solution”. This argument is as fallacious as claiming that trying to avoid political and social problems in a new government via careful constitutional design is “law fundamentalist”. The constitution of a country has deep political, social, economic, and cultural ramifications for said country; similarly, the design and organization of a social network has ramifications for its culture, governance, and overall feel. (Incidentally, this can also be said of less fundamental things on social network, like features and even user interfaces, but that’s a topic for another day).

Mr. Troy is also implicitly dismissive of “restoring voices that have previously been silenced”, presumably assuming people who talk about this are simply referring to trolls and other assholes. While I am sure Musk thinks of the term this way, the truth is in fact more complicated: there are issues with voices that have been unfairly silenced, and I do not mean ex-presidents and conspiracy-peddling grifters with supplement shops. There are many cases of people being banned or suspended due to poor decision making by AI moderation, which is often required on sites like Twitter due to their sheer size. This is also especially pernicious on even larger centralized sites like YouTube, where automated moderation is common, human support is hard to reach, and the general solution to a bad call by an automated moderation tool is to reach out on other social media to the official support accounts and pray that an actual human looks at your case. This is often challenging for existing, established users of YouTube with huge followings; imagine how bad it must be for marginalized users with little to no social presence! With DFOS social media, the need for AI moderation at this scale is eliminated, as instead of relying on centralized moderation of massive amounts of data, users and communities can moderate themselves and interact (or not interact) with others as they see fit.

As for monetization: it is still not clear what model if any Bluesky (currently a non-profit) will use, so the assertions about this that Mr. Troy makes are premature at best. Furthermore, Twitter’s social graph data isn’t going to be as useful as one would think. Most users are anonymous and this graph data is going to be woefully out of date by the time Bluesky launches, given that Twitter will probably be dead at the rate Musk is ruining it as of this writing. As for ActivityPub, it’s ad-free. Smaller instances are often run by small groups or even single people, and larger instances are usually either run by volunteers or are funded via donations/patron systems. For those curious as to whether ad-free cost-free services on the Internet can even exist, I refer you to the ubiquitous one known as email.

Now some will ask “What about harmful but legal content, like disinformation and hate speech? Wouldn’t that proliferate?” From what I have seen, no. While it is true that with DFOS, it is impossible to block a person from the network entirely via technical means, two other consequences arise. The first is that content on decentralized networks is usually spread by human action, and thus content that is widely recognized as offensive generally spreads more slowly. The second is is that, because users and communities are able to moderate and defederate with one another freely, users or communities that spread don’t handle hate speech, spam, or whatever else people don’t like can be blocked off by collective action to the point that they never get the full attention they would on a centralized network.

(For those who think disinformation or hate speech are not protected by the First Amendment in the US, see here).

Part 3: Enter Putin.

Q: What does Dorsey mean, “I trust [Musk’s] mission to extend the light of consciousness?”

A: This is a reference to “longtermism,” the heavily marketed philosophy being promoted by Musk and his friend William MacAskill that asserts the only thing that matters is humanity’s future in space, and that the only goal of the living is to maximize the number of future humans alive, as well as the number of artificial intelligence instances that could possibly exist in the future. This mandate is most often used to brush aside calls for improving conditions and alleviating suffering among the living here on Earth now. Because, the theory goes, giving a poor person a blanket isn’t likely to be as useful for the future of humanity as building a rocket to Mars.

This is wrong, and it is hard to see how this wasn’t deliberate on Mr. Troy’s part. Dorsey’s line in fact refers to another tweet in the very same thread Mr. Troy is discussing, penultimate to the one Mr. Troy embeds in his article.

I love Twitter. Twitter is the closest thing we have to a global consciousness.

— jack (@jack) April 26, 2022

A nitpick with respect to longtermism: while it is indeed a deeply flawed and horrifically problematic philosophy, the description here is rather superficial. A better, more succint description is that longtermism treats a potential person in the future, no matter how distant, as having the same value as a person living now. The issue of space travel is secondary, and from what I can tell, AI intelligences don’t factor into the philosophy (indeed, they are generally seen as something to avoid creating by longtermists). The observation about longtermism discounting present-day human suffering is correct, at least for the version Musk subscribes to.

Longtermism is heavily influenced by “Russian Cosmism” and is also directly adjacent to “Effective Altruism.” Musk’s stated mission, which he intends to fulfill in his lifetime, is to “make humanity a multiplanetary species.” The anti-democratic urge in longtermism is rooted in the belief that “mob rule” will lead to nuclear annihilation; we should, Musk thinks, be guided by “wiser” minds — like his and Putin’s apparently.

This is entirely unsubstantiated by Mr. Troy. From the evidence presented, Russian Cosmism and longtermism have no links; the book Troy links to certainly doesn’t have a mention of any link, or even a single use of the word “longtermism”. Longtermism is a variation (and in my view, very much a corruption) of ethical altruism; the latter more generally seeks to maximize one’s positive impact on the world via the use of evidence and reason, and does not necessarily have the same deep issues of longtermism in valuing the distant unborn over the present.

Since Mr. Troy has presented no evidence for the link between Russian Cosmism and longtermism, we can, by Hitchen’s law, dismiss it without evidence.

Q: Okay, so Musk and Dorsey are collaborators in some weird Mars cult, and don’t actually care about making money from Twitter?

A: Yeah, you’re getting it. Musk also seems pretty interested in helping advance Putin’s “multipolar world order,” which is why he plays footsie with QAnon and MAGA accounts, and pals around with Putin and Dmitry Medvedev. “How are things going in Bakhmut?,” Musk asked Medvedev.

This isn’t that far-fetched, but still requires one to assume that Musk is a knowing accomplice to Putin rather than a useful idiot, which is also an equally satisfactory explanation.

The same cannot be said of the next question-response.

Q: So this is why Musk has been so supportive of the Russians lately? I thought he was on Ukraine’s side, since he helped them out with Starlink?

A: It’s best to look at Musk as a kind of power broker between multiple parties, including Ukraine and Russia. SpaceX was made possible by laws passed by Dana Rohrabacher, Putin’s favorite Congressman. He really doesn’t like the US government, the SEC, and our politicians, and all things considered, he’d probably prefer to replace all of it with an app — call it “X,” the everything app. Elon might say: “Do we really need a government, or could everyone just vote with an app? Be honest.”

For an actual, sensible answer to this question, watch this video by Phil Mason (aka thunderf00t), a chemist and longtime critic of Musk, on the matter:

Personally, I suspect that Russia and Ukraine would be surprised to know they, or any other country in Europe, considers him a “power broker” of any kind. However, this is just a hunch on my part. Feel free to ignore it.

While Dana Rohrabacher is rather cozy with Putin, the legislation he passed was passed in 2004 and essentially opened the way private investment in space travel and exploration. As such, it benefited a variety of private parties other than Musk, not all of whom are necessarily pro-Putin. In any case, implying that Putin was responsible for passing this bill explicitly to help Musk in 2004 seems incredibly far-fetched, assigning a prescience to Putin that seems impressive to a superhuman degree.

While Musk has stated that he wants to turn Twitter into an “everything app” in the manner of the Chinese app WeChat, this seems exceedingly unlikely to happen as I write this given how rapidly Twitter is falling apart under Musk’s ownership.

Q: So Musk wants to overthrow governments, and Dorsey is helping?

A: Don’t be so dramatic. Everyone knows that the only way to advance society is through technocracy. Musk’s grandfather, Joshua Haldeman, was involved in the technocracy movement in Canada in the 1930’s and was arrested — because it was seen as a threat to the government. The philosopher James Burnham wrote in his 1941 book, “The Managerial Revolution” that society should be run by technical managers, not elected politicians. Russian “methodologists” and “political technologists” concluded the same thing in the 1990’s when they realized that “democracy” would just lead to mob-rule by populists. Democracy, they concluded, must be managed. This is what Putin has concluded, also.

So because Musk’s grandfather was a member of the Technocracy movement, that means Musk must be too. Again, this dubious deduction underlies this entire argument, as does the earlier claim regarding Dorsey and Musk being ideologically linked, as does the false premise that Dorsey owns Bluesky.

Troy’s reasoning really is very threadbare, isn’t it? I’m not even going through the rest of the claims. They may be true or false, but they don’t matter given that this sub-argument fell apart so quickly.

Q: So Putin, Musk, and Dorsey share the same vision?

A: It’s complicated. All seem to think a “multipolar world” is a good thing, because after all, shouldn’t Russia get to do its thing and not be bothered by anyone else? That’s “free speech” and opposes “cancel culture,” right? So yeah, that’s aligned with Putin. But Putin himself doesn’t support free speech; his government censors wildly, but it does support speech that breaks the hegemony of the Western elites. As do Musk and friends. This is internally inconsistent.

Yes, Putin is a hypocrite and this description of him and his attitudes towards speech is relatively accurate. While not addressed in this article, Musk is also very much a hypocrite on the matter of free speech, as he has retaliated against employees and even customers for speaking critically of Tesla vehicles in any way. None of this matters because the original linking of Dorsey and Bluesky to all of this is completely spurious, so this part falls apart as well.

Part 4: Crypto Conspiracy.

Pretty much all of this is aging like fine milk as of this writing.

Q: Back to the money thing… won’t advertisers balk at all this potential chaos, and won’t Twitter’s business model suffer?

A: Musk and the people backing all this are more interested in reshaping the global order than in earning fake “fiat currency.” Their real goal is to usher in “hard currency” and re-base global currencies around scarcity and physical assets. So no it really doesn’t matter much what happens to Twitter’s ad model in the meantime. It will probably do alright, and they can probably find other ways to make money, like adding in payments and weird Dogecoin schemes.

Musk is currently planning to charge users $8/month for the verified checkmark, and he’s only going to be accepting real money from what I can tell. This is precisely because Musk needs to make “fiat” money, and given the decline in advertiser revenue, his need will only get worse. Given the current state of crypto prices and the very small number of crypto holders in the world willing to spend money on a 280 character limited cesspit, adding in crypto payments seems extremely unlikely to make Musk enough money to cover the interest on the debt he racked up for the deal.

Everything else here can be solved with another application of Hitchen’s law, because there is no evidence given in links for any of these assertions.

Q: How is Twitter going to help them kill off fiat currency? You mean like replace the dollar as the world’s reserve currency?

(Short answer: it won’t, and they won’t even try, because this whole thing is ridiculous.)

A: As the Russian “methodologists” will tell you, it’s incredibly important to control the information space if you want to alter the real world. As the world plunges deeper into war and economic chaos in 2023–2024, there are real (perceived) opportunities to advance cryptocurrencies and asset-backed tokens to replace the dollar. Whether any of this is realistic or not remains to be seen, but this is what they’d like to pursue — Musk and his deal backers in particular. [Update: Saudi Arabian Prince Alwaleed bin Talal announced he is the second largest shareholder in Twitter; Saudi Arabia has long sought to unseat the dollar as the reserve currency.]

At this rate, with all the stuff Musk is doing, it looks like Twitter will be a dead site months before Bluesky launches, so if Musk is trying to control the information space to alter the real world, it’s going about as poorly as it possibly could. Cryptocurrency prices are still far lower than they were a year ago, there is a high possibility they could sink further if Tether or some other suspected fraud goes belly up, and of course crypto prices are still just as volatile as ever.

Q: So isn’t Musk’s ownership of Twitter a national security risk? Shouldn’t the deal have been stopped on national security grounds?

A: Yes, this is all a national security risk and the deal should probably have been stopped. The fact that it wasn’t is reflective of either fecklessness or capture of the US government.

Techdirt did an article on this. At the time I first read it, I thought it was a little short of the mark, but looking at it again, it’s spot on and a vastly better explanation that Troy’s assertion.

Q: Can Musk and his friends really do all this? Should we really be worried? It seems so far-fetched.

(That’s because it is.)

A: That remains to be seen, and it’s unlikely they can achieve all they’d like to on the timeline they’d prefer. It will likely take years to make Bluesky into a viable product. However, just because it’s hard doesn’t mean they aren’t going to try. As for the broader anti-government agenda, as we’ve seen with January 6th and other attempts to challenge government, even unsuccessful efforts can be incredibly damaging. It’s worth monitoring the direction this takes and, the effects that it has on society and democracy.

At this point, I am not sure about the timeframe Bluesky will need until launch, but there is at least one other competing service, Planetary.io, in the works. I should not that as of this writing, new signups to the Fediverse are continuing at a very elevated rate.

Part 5: Troy’s closing, and mine.

Mr. Troy now gives some run-of-the-mill corrections to some run-of-the-mill comments people make. Of these, only two things are worth commenting on.

God, I am sour after all this.

Q: Boy, Jack really screwed Elon by making him overpay.

A: No, Jack basically asked Elon to buy the company and Elon set the price. …

Really? My impression was that Musk made motions to buy Twitter, Twitter responded with a poison pill, some back and forth happened, Musk gave an obscenely high offer for the company while waiving due diligence, and the company went through with the deal because it was too good to resist for an unprofitable social media company.

Q: Elon never really wanted to do the deal… but they forced him.

A: No, he wanted to do the deal all along, for the reasons I outlined. …

Which is why Musk publicly backed out of the deal after the financial statements came out, publicly disparaged the company, got himself embroiled in a court case in order to force him to buy, relented when he realized he would face deposition, and finally resigned himself to buying Twitter? Did Jack ask him to do that too? Did Mr. Troy just forget to include the evidence or write about it?

For the record I think Elon is a sociopath, and that this all is going to end in disaster — I just couldn’t stand seeing so much shallow, poorly informed analysis proliferate. So I’ve done my best here to set the record straight.

Shoddy analysis…like the article he just wrote, for reasons I have given above. (Again, to Mr. Troy’s credit, Musk is almost certainly a sociopath.)

Mr. Troy closes by hawking both a white paper and another article of his. I think I’ll pass on reading those. I’ve had enough for today.