If you haven’t been on the receiving end of an unsolicited explanation of Bitcoin and other cryptocurrencies yet, consider yourself lucky. The explanations are often obscure, overly pedantic and at the end of it all, you’re still not entirely sure what a distributed ledger is anyway. But what if I told you there was a way to learn nothing about Bitcoin without losing an hour of your life to some crypto jabroni?
On Tuesday, Botnik Studios—a group of writers and artists that put machine learning to creative uses—released a short Bitcoin explainer generated by feeding a predictive keyboard dozens of actual Bitcoin explainers. These explainers served as a training set for the keyboard, which then predicts the next word in a sentence written by Botnik creators based on pattern analysis from the training set.
If you have a newish smart phone, it probably has a predictive keyboard that learns to predict which words you’re going to type based on what you’ve typed before. While this is a good and helpful idea in theory, predictive keyboards often make bizarre suggestions that can lead to hilarious misunderstandings. The Bitcoin explainer produced by Botnik Studios is no different.
“As with everything we write using the predictive keyboard, editing is the whole point,” Jamie Brew, a co-founder and CEO of Botnik, told me in an email. “The writers are continually choosing from the word options the keyboard supplies. Our piano does not play itself!”
This isn’t Botnik’s first trip around the block. The group has also produced lewd versions of Harry Potter and a script for the final season of Game of Thrones, all of which are created using the Botnik predictive keyboard app. You can try it for yourself on Botnik’s website using source material ranging from the Communist Manifesto to Beyonce.
According Brew, the writers for the Bitcoin explainer—which included comedian Zack Bornstein and former Colbert Show writer Rob Dubbin—tried several different versions of the predictive keyboard based on positive articles about Bitcoin, negative articles about Bitcoin, and Bitcoin ads. The final result is an amalgam of these different sources, and it shows in the explainer video.
“To understand how Bitcoin transactions are created, randomly pick a number between one and 30,000,” the video explains. “Now spend that amount of money on Ethereum. This is known as hashing the code to get some of that Bitcoin.”
It only gets weirder from there, but in fairness the AI-generated video still isn’t the most convoluted explanation of cryptocurrencies I’ve ever seen.