Heroic verse: a poetic push for maths and tech
Poetry is humanity’s oldest linguistic art form – and surprisingly well-suited to improving the accuracy of cutting-edge digital technology.
The big issues
Poetry and mathematics may not immediately seem like natural bedfellows, but practitioners of both disciplines can learn much from each other, particularly in a digital world that’s increasingly augmented by artificial intelligence (AI).
For example, machine translation – including the kind that powers the online tools many of us use to translate phrases or websites – relies on statistical analyses of linguistic features, such as common collocations used by native speakers, to make the most accurate predictions. But poetry is a form for people of all backgrounds to play with the possibilities of language and break established conventions. It’s therefore uniquely placed to help machines move beyond the standard constructions they’re currently capable of and into more diverse iterations of human patterns of language use. That would be a real step-change in capability.
Building on previous research collaboration between poets and mathematicians at the University of Bristol, our project aims to investigate how poetry and maths can inform one another, generating innovative insights into AI, engineering and mathematical language modelling.
We’ve set up poetrishy.org, a multilingual and multidisciplinary online lab space where experimental poetry can drive new ideas in natural language processing. Twice yearly, Poetrishy will publish the results of interdisciplinary experiments – drawing from those working in mathematics, technology, engineering, human-computer interaction, digital literature and poetics – in the form of poems, digital productions, mathematical models and more.
We hope to develop a toolkit for interdisciplinary experimentation and provide opportunities for researchers and practitioners from diverse fields to radically improve technology’s relationship to understand human language.
We also intend for Poetrishy to establish and cultivate a network for novel cross-disciplinary experimentation, research and development. At present, there’s no focal point for such a network. By creating one, we aim to lay the groundwork that will influence diverse fields in the years to come and grow new collaborations capable of responding to the complex digital challenges of the 21st century.
- Dr Rebecca Kosick
- Dr Mauro Fazion Filho
- Johanna Darque
- Russell Britton
- Dr Ademir Demarchi
- Dr Andrés Anwandter
- Rowan Evans
Want to find out what we’ve been up to in the past year?
Download the BDFI Impact Report (PDF, 7,156kB)