network effect

Tyler McIntosh

bryophyte
dust grains
in the wind

               token embedding
when did I learn
to speak?

siren wail cuts the night tunnelling in phloem

               return to sender:
deceased
instance spin-up

mother’s hands
in the soil
signals of home
               Turing test
               turning to cry
                              help
first sunbeam
on a seedling
morning prayer

               the model’s curve
               towards a kiss
> break()

The Science

This poem is inspired by research on forests, and the advent of large language models (LLMs). As we adapt to a new age of generative information computer scientists are seeking to measure and define the meaning of consciousness in Artificial General Intelligence (AGI). Concurrently, ecologists have revealed that trees share nutrients and signals through networks of roots and symbiotic fungi.

This poem explores how both evolved ecosystems and LLM exhibit complex interactions and processing deriving from patterns, response to external stimuli, and extraordinarily large networks. It also ponders the importance of interaction between individuals for evolution, questioning where consciousness begins.

The Poet

Tyler McIntosh is an environmental data scientist, geographer, and ecologist pursuing a PhD at the University of Colorado, Boulder. He uses data-intensive methods to explore complex interactions between forested ecosystems, disturbance agents, and humans in the context of global change. Tyler is a published poet who most often writes in forms such as haiku, senryu, haibun, and linked verse. His work has appeared in leading form-specific print and online journals such as Modern Haiku, The Heron’s Nest, Frogpond, Drifting Sands, and Contemporary Haibun Online. His work has been nominated for various awards and included in anthologies of notable contemporary haiku.

Next poem: No. 65 by Echo Guernsey