I haven’t updated a damn thing on my reading list. Why? Well, there are a couple reasons for that.
#1 is that The Selfish Gene is fairly long. And I’m enjoying it so god damn much, I am really spending a lot more time rereading and thinking about what Dawkins is saying than I do in most books. A big part of that is also definitely the fact that it is very new to me. I always sucked ass at biology in school. To me, biology was one of the subjects that was all about memorization. And in high school, that is pretty much true. But wow, this book is so conceptual that it makes me want to scream.
Why didn’t I learn about biology like that in school? Anyways. I’ll save that for the post I make about that book. The real real reason I haven’t updated my reading list is that I’ve spent HOURS AND HOURS AND HOURS on BITCLOUT.
What is BitClout?
I’m not here to explain that to you, they’ve done it themselves:
But boy oh boy. To a data scientist, a social media that I can have full access to all the data? A wet dream… or so it would have been before BitClout actually came out.
The platform has tons of issues. Some people on twitter say the blockchain is complete and utter dogshit. I don’t know enough about blockchains to even have an opinion.
95% of the posts read like your average garbage LinkedIn content but even worse because now there is a monetary incentive to get followers.
But regardless, there have been so many social media alternatives that have had varying degrees of success, and the fact is, BitClout allows anyone to build social media (or more if you think far enough outside the box) on top of the BitClout blockchain with a built in, unique payment mechanism.
As of this second, Friday, May 21 2021 1:43 a.m., this is what my little profile blurb looks like. It looks like any other social media but with a couple additions – the information about my creator coin.
Basically, anybody can trade their $BitClout for specific creators creator coins. Creator coins follow a set $BitClout to $CreatorCoin distribution as shown below:
So creator coins can get created or destroyed as people trade in their $BitClout for them or sell the creator coin back for $BitClout. The more coins get minted, the more the price goes up. A given amount of minted creator coins will always go back to the same amount of $BitClout, so it is really a matter of when you buy or sell someone’s token.
HEADS UP if you plan on making an account, set your founder reward percentage to 100% at the start, or else you’ll have a bunch of people buy your first few tokens when they’re extremely cheap, which means you’ll have to spend a lot more money to get your own tokens later on down the line.
But yeah. I looked into a ton of shit on BitClout and tried some things. Here is what I’ve done so far:
My BitClout Doings
- I attempted to make a bot that buys people’s coins immediately after their accounts are created. I was damn close, but somebody else’s bot was faster. I couldn’t figure out how to sign transactions using the API endpoints. I refuse to believe the other guy’s bot used selenium, because my selenium implementation just was not cutting it.
BUT, honestly I think the dude wasted all his money. The % of new accounts that were fake (and likely made to get a few cents from the coin-buying-bots every time) went up MASSIVELY in the last couple weeks. I bet the guy buying these coins will get a pretty shit return overall. Or maybe I’m just salty that he beat me and he’s hooping now. No way to really know.
- That brings me to my second point: a TON of the new profiles are spam accounts. I checked this all sorts of different ways and every method I tried I consistently got 50-60% of accounts were spam. My favorite method to test whether accounts were spam or not was to check the profile pictures used. When you create a profile, you aren’t able to select a default picture, and you have to choose a profile picture.
So when people made new accounts, I would check to see how many times their exact, pixel for pixel, profile picture had been used in the past. I hand checked a bunch to see the threshold at which point it seemed like the accounts were for sure not real. Even at 3 copies of the same exact picture, most of the accounts looked 100% fake, and so that was my criteria. There were several profile pictures that had been used thousands of times each. I checked them, and saw that the guy with the coin-buying bot had bought all of their tokens (at least all of the ones I sampled). Poor guy. Maybe you were faster than me because you didn’t check these simple things.
- I’ve been learning Go and to some extent TypeScript to keep up with the core code. I couldn’t read a single line of it in the beginning. Now some of it makes sense! I’ve also been going through some very basic blockchain implementations to get a better sense of how consensus actually works. Maybe one day I’ll be a goated blockchain dev. Probably not but lol who knows
If you decide to make a bitclout or are already on, HMU! bitclout.com/u/vinnymeller