using that variational autoencoder I trained on 1m lines of poetry—making a poetic form I'm going to call the "variational sestinoid," which forms its stanzas from samples in progressively larger circles centered on the same point in the latent space

(I like the way these sort of unfurl, elaborating on the same spot of the poem-space, producing variations that retain lexical and semantic links to each previous stanza)

@aparrish this reminds me of an effect i used to try for by chaining successive variations and recombinations of small poems.

i don't think it was ever a very _successful_ thing - like i don't think it really put anything across to my readers other than confusion about what i was doing, and honestly it probably just made the individual poems worse - but it was interesting from inside the space of my head, at any rate.

@brennen the wonderful thing about being an experimental poet is that "interesting from inside the space of my head" is good enough!

@aparrish indeed!

i am, to the extent that i'm a poet at all, utterly conventional, but one of these days i'm going to have to devote some energy to text generation just for that very reason.

@aparrish dammit even the poets are going to end up unemployed due to automation! (This is pretty impressive. But also scary. But mostly impressive.)

Sign in to participate in the conversation
Friend Camp

Hometown is adapted from Mastodon, a decentralized social network with no ads, no corporate surveillance, and ethical design.