- Download a website or torrent of song lyrics
- Parse the result (with BeautifulSoup)
- Use NLTK (Natural Language Toolkit) to lookup Wordnet synsets (homonyms, holonyms, hypernyms, antonyms, meronyms, synonyms and lemmas) as replacement candidates.
- Replace all the words in the poem (retaining the words-per-line and verse and grammar of prepositions and pronouns)
- Sift through ‘the reservoir‘ of words that returned no results from synset search to find dense tweets worthy of a bot.
- Or check out the [ for now, terrible] 5,067 poems created in 8898 secs
Have a peek at a few results posted on my twitter feed:
The poetic problem is that these poems lack insight. Blastocytes have orders of magnitude more experiential insight than these poems.
The poetic problem that arises also motivates the appeal of these poems conceptually. At a linguistic level, code-salad-bot-poems extend the project initiated by LANGUAGE poets to shift poetry’s emphasis away from lyrical transcendence to contextual linguistics.
Density obliterates the reading subject’s potential for narrative engagement even as the formal syntax invokes echoes of such narratives. Integrity and perception are replaced by a pure implacable vertigo.
The fact that these poems can be computationally generated at extremely rapid rates obliterates the dexterity of author as collision operator smashing together words at increasing velocities to see what stray numinous trace particle paths might etch on meaning’s sensor.
The author is neither me nor the original poets who provide the rhythmic and structural templates, but the momentum of culture itself, an aggregate of words protuding out of mouths, filtered thru conversations, social network servers and random cultural trajectories. The author resurrects, readers die under an overload.