Fallout Followup

A day or so after I published my article about griefing the "Crypto Deleted Tweets" account, I was contacted by someone (who shall remain anonymous) who is connected to Twitter. They informed me that they read the blog post and while my bot idea was an interesting response, the Crypto Deleted Tweets bot was in violation of Twitter's Developer Policies and that I should expect the account to be taken down. A day later, the account was gone, at which time I posted this:

The ensuing mob of upset people all jumped to the conclusion that I had instigated an "army" of Twitter users to mass report the bot into oblivion for "harassment" because I was "upset with the bot." Not only was the mob wrong, nobody even bothered to ask me if I knew why the bot had been suspended!

Source: https://twitter.com/udiWertheimer/status/1205070189655740417

Narrator's voice: "There wasn't. The one person claiming responsibility was bullshitting."

Apparently this is a well known scammy account who has beef with the Deleted Tweets bot. Source: https://twitter.com/bitrandy/status/1205380247505362950

A lot of people were understandably upset.

Source: https://twitter.com/MartialCrypto/status/1204774376383664131

I received more replies like this than I could count, many were far more vitriolic. Interestingly, about 99% of them came from anonymous accounts! Very few of these tweets made it through my filters to notify me, so it was pretty hard to tell if the outrage was a sybil attack or if I had just managed to tickle some subgraph of Crypto Twitter with which I'm unfamiliar. Oddly enough, I later noticed that several of the more vitriolic accounts also got suspended, which made me even more suspicious...

The Value of Crypto Deleted Tweets

From the outcry over the account's suspension it became clear that there were several types of folks who found value in the information provided by this account.

  • Crypto traders who want to catch scammers erasing poor predictions
Source: https://twitter.com/thecryptomonk/status/1206941862835183616
  • People who want to know which accounts are making bogus claims and then going back on their word and trying to preserve their reputation by deleting the evidence. Similar to the trader predictions, but more generic.
  • Folks who use schadenfreude to troll (can't say I care about this use case)
  • Folks who delight in seeing cringe posts (ditto)

Terms of Service Violations

The real reason that the account of the bot(s) and of the bot author were removed by Twitter is because the author was flagrantly and repeatedly violating several terms of using Twitter's APIs, such as not honoring user blocks.

Source: https://twitter.com/CryptoDeleted/status/1183815911691685891

This is a violation of section I. B. 5. of the Twitter Developer Policy which states:

Do not exceed or circumvent limitations on access, calls, sharing, privacy settings, or use permitted in this Policy, or as otherwise set forth on the Developer Site, or communicated to you by Twitter.

The bot was also in violation of section I. C. 3. of the Twitter Developer Policy, regarding "Respecting Users’ Control and Privacy" which states:

If Twitter Content is deleted, gains protected status, or is otherwise suspended, withheld, modified, or removed from the Twitter Service (including removal of location information), you will make all reasonable efforts to delete or modify such Twitter Content (as applicable) as soon as reasonably possible...

Clearly there's some room for interpretation with this clause because it says "make reasonable efforts to remove data." I can only assume that when Twitter observed the bot intentionally doing the exact opposite, they considered that a clear violation.

Kudos to the ONE person on Twitter who actually got it right!

Source: https://twitter.com/odtorson/status/1204728594771193856

You are, of course, free to dislike the above terms, but anyone who generates API keys for a Twitter app has to agree to them. The owner of the deleted tweets bot broke their agreement with Twitter, so Twitter removed their account. That's the TL;DR of this saga.

FAQ

To clear up some other common questions and avoid repeating myself a hundred times:

Q: Why are you so afraid of having your deleted tweets broadcast?
A: As I stated in my first blog post, I'm not. Rather, I realized that once the bot started listening to my account it had subscribed me to a new "game." I decided to play that game... adversarially. You're welcome to disagree with my tactics; it's not called "griefing" because it's polite.

Q: Can't you see the value provided by the Deleted Tweets service?
A: As I stated in my first blog post, yes. I never claimed the concept has no merit.

Q: Why are you whining so much?
A: If the tone of my posts comes across as whiny, that probably reflects more on you than me. I'm making an effort to be transparent even though I have no obligation to do so.

Q: How dare you celebrate censorship, you hypocrite?
A: My goal was to use my own speech to create noise that incentivized the bot owner into improving their filtering of uninteresting deleted tweets. That's not censorship by any means; it's basically the opposite - information overload. I started playing a new game and didn't even know all of the rules at the time; I will admit that it was somewhat amusing to learn that the bot was violating the rules of its own game.

Q: How can you call yourself a Cypherpunk, I've lost so much respect for you!
A: Cypherpunks write code... which is what I did with my griefbot. Cypherpunks are generally opposed to censorship and monitoring by government agencies, which I am. Twitter is a private platform that is not censorship resistant, which is a shame, but that's the reality of the situation - don't hate the player, hate the game.

Q: Why not just create your own deleted tweets bot instead of being an asshole?
A: Because this was a lot less effort, more entertaining, and it later turned out that writing my own deleted tweets bot would have violated Twitter's terms of service and been similarly shut down.

What Now?

It was only a matter of time before the deleted tweets account got suspended; in hindsight it's amazing that it operated for years without receiving any scrutiny. Did my blog post cause the account's demise to come earlier rather than later? Sure, I'll take responsibility for that; hopefully you can also realize that the bot owner made a choice to add me to the short list of accounts it follows - by doing so it attracted my attention and I put a spotlight on it. I believe that the best way to improve systems is via scrutiny and stress, so you'll get no apology from me for my response.

I don't expect the bot author will succeed in bringing this account back onto Twitter, since Twitter will shut down any API key that they discover being used to violate the Developer Policy. Rather, the bot will be relegated to posting to alternative platforms.

Ultimately, information wants to be free, and Twitter can't control the flow of information. As such, I'm quite sure that the bot will continue to operate in another form. The game is afoot!