The Twitter Files: The Lost Chapters
The Twitter Files exposed Orwellian levels of control at Twitter, but some things were left out.
This article has gone through several rewrites. It was originally going to be a long exposé on the many of ways Twitter had become a corrupt, untrustworthy organization that was involved in an incestuous relationship with the Federal Government, politically biased media outlets, and political organizations.
Many items I had intended to write about have now been covered by the release of what has been called The Twitter Files. Independent journalists chosen by Twitter’s new owner, Elon Musk, were given direct access to internal documents and communications — giving them the ability to go into great detail the internal motivations and reasoning behind decisions the company was making.
As of December 31th, there have been 10 installments under the title of The Twitter Files
How and Why Twitter Blocked the Hunter Biden Laptop Story
Supplemental - Former FBI agent secretly sanitized documents.
The FBI Treating Twitter as a Subsidiary
Supplemental - FBI pushes Twitter to agree with it about foreign actors
How Twitter Quietly Aided the Pentagon’s Covert Online PsyOp Campaign
Twitter and ‘Other Government Agencies’ (e.g. the CIA)
The Lost Chapters
The Twitter Files are still being released over time, but there are at least two very important items that I believe are unlikely to be included. These items are still very important to the whole story.
Quote Tweets Only
On October 9th 2020, 25 days before the Presidential Election and five days before the Hunter Biden Laptop story was first posted by the New York Post, Twitter made a radical change to the way users interact with tweets on the platform: They modified the retweet functionality of twitter.
Prior to October 9th 2020, when a user wanted to retweet something, they would click the retweet button 🔃 they would be given a menu with two distinct options.
A normal Retweet simply shares the original tweet, with no additional comments or frills, to all a user’s followers. A Quote Tweet, however, lets the user create a new tweet of their own with comments, and places the quoted tweet inside.
On October 9th, Twitter announced that they were removing the regular Retweet option from the user interface.
A regular user of twitter was no longer provided the option of simply retweeting a tweet. The user interface only prepared a Quote Tweet for the user to compose.
The change caused lots of confusion with most users unaware of the announcement, only discovering the change when trying to retweet. Some users discovered it on Oct 9th, when Twitter did a partial rollout of this change
And others discovered it on Oct 20th, when the change was made global.
Controlling what users think
The articulated justification for this change was that it “encourages everyone to consider why they are amplifying a Tweet, and brings more thoughts, reactions & perspectives to the conversation.” In isolation that would seem like a plausible motivation for the change. There are three important facts that cast extreme doubt on, if not completely nullify, that idea.
First, the stated goal of the change was not something for which Twitter had evidence would likely result. They had not conducted tests, prior to deciding to make the global change to the platform, to see if it indeed encouraged “more thoughts, reactions & perspectives to the conversation.” If the goal was genuine, the election itself was the experiment.
Second, the change was announced in a tweet thread that was all about the myriad of ways Twitter intended to control conversation on the platform, not about how to encourage more perspectives. This change would have been the only circumstance where it was argued by Twitter that increasing the number of different perspectives was useful in helping fight what Twitter considered “misinformation.”
Third, if Twitter had high confidence its stated goals would be met by the change, and those goals were to encourage “more thoughts, reactions & perspectives to the conversation” why would the changes explicitly be called temporary? “We’ll temporarily ask people to add their own commentary before amplifying content by prompting Quote Tweets instead of Retweets” is what the announcement said. Why would someone intend only to encourage “more thoughts, reactions & perspectives to the conversation” for a short period of time?
As we will see, Twitter did not view adding “more thoughts, reactions & perspectives to the conversation” as a positive, and that was not their actual goal.
The real impact
While majority of users were unaware of the announcements, they had no choice but to notice the changes to Retweet behavior through the Twitter user interface. Confusion, shown above, was the initial reaction until the changes became “just the way twitter works now.”
When users would press the retweet icon 🔃, they were now presented with a prepared Quote Tweet, and their cursor blinking with “Add a comment” on screen.
Users who did not actually want to add comments quickly circumvented the demands by adding variations of “nothing” to satisfy the instructions of the prompt. Single items of punctuation such as a period (.), or one word comments, were exceptionally common ways of dealing with the change.
Both “real” and “workaround” comments had a significant impact on how information was shared on twitter: it reduced it.
Tootsie Pop Tweets
The forcing of Quote Tweets increased a phenomenon that I will call “tootsie pop tweets.” Most people who are at least of the millennial generation will be familiar with an old classic commercial for Tootsie Pop, a lollipop with a tootsie roll in the center. The classic commercial features a cartoon child asking everyone they meet “how many licks does it take to get to the center of a Tootsie Pop?” Nobody knows because most people give up part of the way through.
The impact of the Quote-Tweet-only change was to obscure the original content inside of an increasing number of layers of Quote Tweets. How many clicks does it take to get to the center of a Tootsie Pop Tweet? Nobody knows because people give up part of the way through.
Imagine an article were written on Substack by your favorite author. Maybe even this author. You read the article and post a link to it on twitter saying: “Everyone needs to know this!” Someone who follows you clicks the article and reads it. They click back to return to the original tweet and. They are prompted to add a comment. They say something like “I agree!” and they send the quote tweet.
Someone that follows them clicks through to read the article. When they’re done the click back a few times to return to their timeline where they began. They hit retweet and are asked to add a comment. They write “wow!” and send the quote tweet. Their followers see "wow!” with a quoted tweet of “I agree!” At some point, the original tweet is nested inside many quote tweets, and the central topic of discussion is buried. The content’s spread is suppressed.
This used to be a playful thing people on twitter would sometimes do as a gag, but now it was the way Twitter worked by default. For an example, consider this Quote Tweet
See how many clicks it takes you to get to the center of that tootsie pop tweet, and ask yourself how many times you’d bother to do this while using twitter in a regular day, in order to see if the original tweet was something you were interested in.
Covert disclosure
If you’re with me this far, then OK, Twitter changed basic functionality of the platform which made an occasional gag (Tootsie Pop Tweets) the standard behavior of the platform, but this would impact everyone equally, so it is a wash overall. Perhaps not.
Some users with a history of trying to make computers do whatever they want, like your’s truly, discovered that if you did not enter anything into the quote tweet section, it would become a regular retweet.
Tech outlets were given the press release version of the announcement which contained a single sentence, a little secret, not mentioned in the tweet thread: you could get around the Quote Tweet feature altogether by leaving it blank
From the press release version:
Starting today, when you click or tap the retweet icon, Twitter will pull up the Quote Tweet composer to encourage you to write something about that tweet before you share it. You don’t have to write anything if you don’t want — just leave the composer blank and hit the retweet button to retweet like you normally would.
This information was technically disclosed in many locations by media who used the press release to write their story. Those people could tell their friends how to circumvent this new behavior.
But this important detail was not disclosed clearly on the platform itself. It was intentionally not disclosed in order to nudge users to use Quote Tweets instead of Retweets (or to not use the features at all, another option Twitter liked).
It’s all about control
This Quote Tweet change had its negative effect on information sharing itself. But it was also not operating in a vacuum. It was part of a multitude of changes whose combination was a powerful tool in shaping user’s thoughts on what was happening in the world.
Well followed mainstream news outlets post their stories without trouble. If Twitter wanted to, they could slap a notification on the tweet of the story to add their own opinion as being more official than the story itself.
Twitter would do so, however, by using the mainstream outlets to determine whether content on twitter deserved such a label. The Twitter Files show content moderation teams actively googling tweets they read to see if they could get away with labeling something as disputed or misinformation.
Mainstream outlets thus dictated the moderation team’s interpretation of what was acceptable content on the platform (influenced of course by whether the moderators considered the network itself reputable - the reader can decide for themselves how the moderation team would probably view Fox News). If users read non-mainstream stories, the Quote Tweet feature helped suppress its spread by the intentional addition of interface complexity. Twitter moderators googling mainstream outlets then arbitrated truth for the users sharing the unapproved information.
If, despite the friction and labels, the content became viral enough, the content moderation teams would add their own explanation for the trending item on the trending bar - telling users what conclusion to make about the content prior to ever seeing or hearing about it otherwise.
Failure is success
Twitter’s proclaimed goal for changing the Retweet functionality was in decent contrast to all of the other changes they made and the articulated motivations for those changes. The stated goal was “encourag[ing] everyone to consider why they are amplifying a Tweet, and [bring] more thoughts, reactions & perspectives to the conversation.”
As said earlier, this would be the only situation where Twitter would claim to view more perspectives in the conversation as a positive. Twitter in fact did not view adding more perspectives to the conversation as a positive, because when the election was one week behind us, Twitter claimed the reduction of all retweets was a success.
…on a net basis the overall number of Retweets and Quote Tweets combined decreased by 20%. In short, this change slowed the spread of misleading information by virtue of an overall reduction in the amount of sharing on the service.
Because of the heated electoral disputes at the time, Twitter obviously didn’t remove these controls as they were still of high utility. They made the announcement the retweet functionality would remain broken, from the user perspective, for the time being. They said “we are taking more time to study and fully understand the impact of this change and are leaving it in-place for now.”
Twitter didn’t restore the normal retweet functionality until December 16th, 2020. Taking them at their word and not at my more cynical, but correct, perspective, that means it took them an additional five weeks to discover users were simply adding single words to the Quote Tweets
We hoped this change would encourage thoughtful amplification and also increase the likelihood that people would add their own thoughts, reactions and perspectives to the conversation. However, we observed that prompting Quote Tweets didn’t appear to increase context: 45% of additional Quote Tweets included just a single word and 70% contained less than 25 characters.
Tracking codes
During COVID and through the United States 2020 Presidential Election, Twitter had amassed a fun new set of toys they were eager to play with. Adding “context” to trending topics became a regular feature. Marking tweets as misleading, disputed, and prohibiting sharing or replying, became an everyday part of the twitter experience.
With the COVID vaccines from Pfizer and Moderna (and the red-headed step-child J&J) having a global campaign set on universal vaccination, their new tools were retargeted to allow them to control what users thought about vaccines.
The Biden administration was not satisfied, however. By July, 2021, the White House was not only actively pressuring Twitter to blacklist and ban specific users, it was pushing Twitter and others to have a global user-banning policy: A user banned from one social media platform should be banned by all.
While no global, cross platform, ban policy has yet been officially adopted, the direction of increasing government pressure towards controlling information and individual users on the whole of the internet continued throughout 2021 and 2022.
Unannounced changes
The tracking codes were not added to all accounts on all devices at the same time. I don’t know when it was first turned on for my account (or limited to what versions), but I first noticed the tracking codes being attached when I copied tweet URLs on January 28th 2022. I immediately notified others I talk to in private.
Some accounts saw the feature at least as far back as August 2021
These changes were not announced anywhere I have been able to find. There is no press release explaining the feature or its intended use (again, that I could find). Subsequent to Elon Musk officially owning Twitter, the former CEO Jack Dorsey, directly acknowledged them and suggested he did not find them as being too innocent
In the replies, Jack hints that he may have been against them when he was still CEO, but perhaps overruled by the board of directors
Hopefully future Twitter Files will illuminate this internal discussion.
What are they
Twitter obviously knows a good deal about what you do with your account while you are on twitter. It knows who you follow. It knows who follows you. It knows the tweets you like, and the replies you leave. It knows the people you message inside of twitter. This is information that, one way or another, you are volunteering to them.
But Twitter also exposes that information to others. If you go to their developer portal you can register to use what is called an API (Application Programming Interface) which lets you access much of the data that is available to regular users’ eyeballs, but with the ability to grab more of it faster. Arbitrary users’ followers for example, are available for someone to program putting into a spreadsheet daily, if they wanted.
Those with some decent programming skills can easily work with database systems that are designed to look at network type data intuitively. These systems allow those with such skills to make interactive charts showing who is connected to whom, and how, on twitter.
“Researchers” and those with some nice monetary resources to spend can get versions of access that allow them to pull vast amounts of this data quickly.
Twitter, of course, is able to see who you are following and who is following you in real time. They are also able to see whose content you like, share, and who shares yours, in real time. When you do anything on Twitter, you’re sending information to their systems and often information identifying your account is included. This makes sense.
The tracking codes allow Twitter to collect information you may purposely not want to share with them: your off-Twitter personal connections.
As fully corrupt as Twitter became, knowing users' off-Twitter activity allows for more flexible censorship and deplatforming. In an environment where Twitter is already suppressing information, and is targeting users based on activist employees and government agent whims, collecting hidden connections gives those censorious actors inside Twitter and the government, by proxy, greater ability to focus on targets whom they deem are high value.
If the users are not careful, Twitter could (and currently still can) collect information about users’ external network connections and specifically what information on Twitter is being shared between them. If that information is considered to be a problem to activist employees inside of Twitter or the government, the incestuous relationships between them, the media, and the government enable a wide array of ability to suppress the exchange and harass those sharing.
Consider what could be done with a robust cross-platform suspension policy in place like the White House wants.
Preventing this (on twitter) is simple
The way the code actually works is very simple and getting rid of it fairly trivial, fortunately. Consider a link to one of my tweets:
Look closely to where you see the question mark (?). Then look just a little to the right to see the ampersand (&) and “t=” in red. This is where the tracking code is.
Everything to the left of the question mark is necessary to identify the tweet you are linking to. Everything to the right of the question mark is not necessary. To the right of that question mark, there is a small code in blue, “s=”, to tell twitter if you’re using Web, iPhone, or Android. In this example, it’s 20, Web. The long sequence of seemingly random letters and numbers, LcF7dtFJFk96gs9uHhKNcg, is the actual tracking code.
When you “Copy link to Tweet” or “Share Tweet via …” using the twitter interface, these codes get attached. When you send them to someone else — perhaps through text message, or other chat program — if that person clicks the link, Twitter will know you sent them the link.
Destroying that code in some way (e.g. just deleting it all together) will stop Twitter from associating the sender with the receiver. All you need in order to visit the tweet is the left of the question mark (?).
Constantly deleting the tracking code by hand is rather annoying, however. So here are some instructions on how to use some helper programs to do it for you.
Desktop - Google Chrome
On a computer, if using Google Chrome for a web browser, an extension can be installed called Copy Clipper, which which can modify things you copy for you. Click the link to Copy Clipper, and on the page for the extension click the button Add to Chrome.
Once added, you should see this notification.
If so, locate the Copy Clipper Options window
You should see the following window open in a new tab.
Copy the below intimidating line of code and paste it above the line that says “# Add all new patterns above this line.”
!(https?:\/\/(www.)?twitter.com\/\w+\/status\/\d+)\?(s=\d+)?&?(t=[\w-]+)?!$1!
Press the save button at the bottom and close the tab. Now when you go to twitter and copy a link to a tweet, the tracking code should be gone.
If you need similar instructions for a different browser, send me a message on twitter @drrollergator and I’ll see what I can find for you.
iPhone/iOS
Similar to above, there are ways to add menu options to iPhone/iOS that act as a special copy feature, which can remove this code for you.
I do not have an iPhone to write instructions myself. I did have a friend record steps to demonstrate the functionality in full, to help readers. Tech outlets have written guides for iPhone/iOS for users to be able to clean links across the internet in a generic sense. It is those guides that serve as a basis for the video’s further instruction.
The code my friend used in the “replace” location is this:
(\?)utm[^&]*(?:&utm[^&]*)*&(?=(?!utm[^\s&=]*=)[^\s&=]+=)|\?utm[^&]*(?:&utm[^&]*)*$|&utm[^&]*|www.google.*/amp/s/|/amp
Hopefully, this information can help you set this up on your iPhone/iOS device. If you have trouble let me know on twitter, and maybe we can find someone to help figure it out and perhaps amend these instructions.
Android
I do have Android. A similar application to the iPhone/iOS is available that I have used. The application I am using is called Clean Share.
To my eye, Clean Share appears simpler to get working for this purpose. Once installed from the Google Play Store, opening the program on your Android device gives the options screen.
There is a default rule already set up. This rule only needs one modification. That is to add |t to the end of the rule that is already there, and press DONE. The vertical bar and the letter t are both needed. The vertical bar can be thought of as working like a comma, telling the program that “t”, the next entry, is also something to remove from links copied using Clean Share. “t” is where the twitter tracking code is located.
Once set up, now on twitter on Android when you click the share icon on a tweet, you can click the “Share via…” button.
When you do, you will be given a choice of what program to use. You would select Clean Share from here. Before selecting Clean Share, you will see the tracking code onscreen.
After selecting Clean Share, the screen will probably flash, as the program does its job quickly, and the tracking code will be removed.
You can then press the “Copy URL” button and share the new link safely with whomever you originally wanted.
All of these fixes are a hassle to some degree. These methods are less obnoxious than manually erasing the tracking code every time you wish to share it, but they not free of nuisance. Hopefully Twitter, under Elon Musk’s management, will agree and discontinue the system that creates and adds these tracking codes in the first place.
The pressure will continue
While the Twitter Files have been important in making sure everyone is on the same page about what kind of manipulations and pressures exist in the major social networks, the ultimate outcome is that those who were worried before are still worried, and those who were not worried before have found reasons to argue all the activity was acceptable, and that increasing the controls important.
Elon Musk buying Twitter and allowing everyone this glimpse does not free him from future demands of government, media, and activists to control information flow and what speech is permissible on the platform. Elon has demonstrated poor judgement in some of his instincts in this regard, having recently caused a large controversy by instituting a rule to suspend all accounts that mention or link to a list of other websites.
After a large outcry, Musk did reverse this policy. Ultimately, responding to feedback in such a way is better than the previous management. But it does demonstrate that Musk is very capable of making extremely bad content management decisions and at first be very blind to consequences. More subtle policies whose deleterious effects are more in the aggregate than in the individual cases, may not reach his attention to reconsider in the same ways.
Hopefully readers of the Twitter Files and this article will sustain their concerns over time, and think about the actual impact of rule and platform changes as they are implemented. While not a perfect protector, being actively aware and vocal in disapproval can work to at least slow this continued suppression of political information and speech.