Twitter by the bots for the bots

I often used to tease my son, that it was yesterday’s bread tomorrow, when bread was completely stale. Little did I know that this would happend to my Twitter handle! Anyone following me, knows that my Twitter account @DonTravlos is completely unable to follow any other account. The detailed situation is this: I can follow people but about an hour later I automatically unfollow them! What this means is that my timeline is empty. My follower count on the other hand is actually rising despite the fact that I barely post. Why? Because my timeline is empty, what’s the point in looking at Twitter.

I used to use Twitter to keep up with what was happening in cryptoland, well no more, I now depend, as I have always done on reddit.

Thinking about this Twitter situation, it occured to me that it’s a Sybil problem. Sybil was a 1976 film of a book of the same name. The main character, Sybil, is severely abused by her mad mother, causing her personality to fracture into multiple identities in an attempt to protect herself. In the crypto space a Sybil attack refers to one where one bad actor can appear as many because of the ability to clone nodes or information seamlessly on the Internet. The issue for Twitter is what it can do about bot, scam, fake or fraudulent accounts. Well, a bit of thought will show there’s not a lot they can do short of requiring every account to be linked to a real world user. Then I thought hang on; what if they require a user to have their webcam or the front camera on their phone operational whilst using Twitter. With modern deep learning neural nets they could monitor whether there was an actual person on the other end and if that person was the same one who had submitted ID documentation like a passport.

Don’t get me wrong: I think this is an absolutely terrible idea, but I would not put it past the social media giants like Facebook to try it via a “sorry it’s not us its a government regulation” argument. Just like KYC/AML legislation in the financial sphere.

Besides the fact that I don’t think it will work, it should be pretty easy to put a feed through to Twitter of you working at your desk, while your bot was tweeting avidly away trying to influence the latest election.

Why won’t they do it? Because of the immense amount of work it would entail. Sure the neural networks are fast and can be built into ASICs, but the amount of processing that would need to be done would be huge. A bit like all the of the processing that goes into securing those other permission-less networks, Ethereum and Bitcoin.

So what Twitter is doing is trying to run a permission-less network without doing the “work” to secure it. Why? Because if they do the work to secure it, their profiit will drop to zero or will be negative. If ever you wanted to see a use case for blockchain and crypto economics: microblogging and social media is it.

From what I can see Twitter is just not getting it, I have another account that can actually follow people but the content on that feed has degraded so much that I only check it a few times a day. I am so confident of my thesis that I am actually considering shorting Twitter stock and later Facebook.