This is the consonant convolver. type or paste some text in the box below and it’ll be converted into chat-speak,Â by removing all of the vowels, obviously.Â Then it re-composes a sentence from words with the same chat-speak equivalents. I have personally found both conversions to be very entertaining. Try it out!
First, words and their meaning have a stronger significance for most people than the subject material of the most important algorithms. Google’s pagerank or the High-Frequency Trading (HFT) algorithms are rendered doubly opaque for the arcane nature of the subject material they’re dealing with in the first place.
I also like that it’s obviously doing (admittedly minor) harm to the sentence and the ideas within it. It’s so easy for everyone – especially programmers – to believe that the application of algorithms and software to data is going to result in its improvement. So often the reality is that at best the algorithm does nothing constructive, and at worst it does real harm. To render this harm visible in a field familiar to more people helps shed light on the dubious benefit of some of our machinations over data.
Third, I hope that it helps people less programatically-minded to consider what might be possible with algorithms. While it’s true that this sketch is unlikely to do any good, consider that in order to run it, your browser first takes a list of all words in the English language – some 70,000 of them – and applies the convolution to each. Even despite the stupidity of the operation, one has to be impressed by the speed of the algorithm and the implications of that.
Lastly, I feel like this gives us a glimpse of a really amazing insight. This particular algorithm is obviously silly, and that silliness is obviously mine. While algorithms can draw upon and derive objective truths, that doesn’t remove their bias. The selection of one fact or domain of knowledge over another is centrally important to the design and operation of an algorithm and this goes unnoticed too often by programmers and non-programmers alike.
(I should note that the words come from here. It is an amazing resource to have on-hand!)