Internet bots are software applications (scripts, really) capable of performing an action or generating content automatically or with minimal human intervention. Their existence is neither good or bad per se and yet, they have acquired a fairly negative reputation over the years. Many of the bots internet users tend to encounter (this leaves out the ones that work on a deeper, more structural level, such as Google Crawler), crowd social networks, annoy through repetitive posting, spread false or incendiary information, harass or offend, or simply lurk in cyberspace, taking up room and gathering data for questionable purposes. But, just as there are bad bots, there are those that are necessary for the normal functioning of the internet and studies have shown that even those content-bots (the ones that get obnoxious at times) can be used for positive purposes, provided the humans behind them aim to do so.
Fair enough, but when do we circle back to literature? Right this instant, because on Twitter, where it’s all about words, there are quite a few bots of noble intentions —such as spreading the love for literature and awakening a creative spark in their followers. Here are five of the most interesting ones:
1-Magic Realism Bot (@MagicRealismBot)
The literary bot by @chrisrodley & @yeldora_ is probably the most popular one on Twitter. It’s been programmed to generate microstories that are vaguely reminiscent of narratives written in the style of magic realism. Over time, it has become more cohesive and has incorporated a large number of syntactic structures and interchangeable elements. Borges would have been intrigued.
2-This is Just to Say (@JustToSayBot)
A bot that that exists solely to create versions of that very famous and infinitely parodied poem by William Carlos Williams. They’re amusing, even if finding meaning in most of them is a real challenge.
they were programmed
and so fun
3-Tales of the Arcane (@Lilarcanebot)
This bot combines random fragments from the book The Pictorial Key to the Tarot by poet and occultist E. Waite to create brief descriptive sequences of an esoteric inclination. It was programmed by yours truly in Python, as a very humble tribute to Waite and Ítalo Calvino and is no longer active because I carelessly deleted the script while reformatting my laptop. (The tweets left on the account are pretty nice, tough!)
4- Ulysses Reader (@UlyssesReader)
In their most basic form, Twitter’s literary bots take a text and parse it into segments with the right character count to be posted on the platform. The two bots by Timo Koola consecrated to James Joyce (Ulysses Reader and Finnegans Wake) do just that. There are several others, including Moby Dick at Sea, which publishes quotes from Herman Melville’s novel.
5- Shelley (@shelley_ai)
Ulysses Reader is a predictable corpus-fed bot, while the others on this list are text generators. Shelley is a more sophisticated project. Created by Pinar Yanardag, Iyad Rahwan and Manuel Cebrian from the Scalable Cooperation Group at the MIT’ Media Lab (“Reimagining human cooperation in the age of social media and artificial intelligence”), it combines a multi-layer recurrent neural network and a deep learning algorithm that is enhanced though user participation. Before starting to write on its own, Shelley read and analyzed classic texts and more than 140,000 amateur horror stories posted on the Nosleep subreddit. This is how it works: the bot tweets the beginning of a new story and any user can reply continuing the story with the hashtag #yourturn. Shelly answers, then another human jumps in, then Shelley and so on until they reach #theend. The bot has been inactive for a while now, but both the account and shelley.ai are worth checking out for completed stories.
Some bots make mammoth books seem less daunting, while others pique our curiosity or offer a goal for new readers (finish the book before the bot makes its way though it, for instance). For Adam Hammond, who wrote a simple tutorial on how to code a text-generating script, literary bots partake of a tradition that goes back to the Dadaists and their découpés. And, as regards the threat of artificial intelligence, the aim of all these creations is to inspire and, at best, broaden our understanding of what is and isn’t “proper literature”. Nobody here is actively trying to erase the author or kill the novel (quite the opposite, in fact). Iyad Rahwan aptly summarized the absurdity of that particular concern in this interview: “If we can build machines that understand the very essence of human experience, we would have bigger problems than simply losing jobs in creative writing.“