The CROSS funded Tracery and Chancery project, which focused on computational support for casual creators, officially graduated from the Center’s incubator program after the October 5, 2020, Industrial Advisory Board (IAB) meeting. The project, led by CROSS Incubator Fellow Kate Compton, was recommended for funding by the CROSS IAB in October 2018, and began as an incubator project in Spring 2019 after Dr. Compton successfully defended her PhD dissertation.
Dr. Compton is a long-time generative artist, inventor, and programmer who specializes in building algorithms that create strange and exciting new experiences. Her work aims to design artificial intelligence (AI) to augment human creativity, and to create tools that bring AI into the hands of non-traditional practitioners, including poets, artists, and kids. As part of that work, she created Tracery as a class project at UCSC and then open sourced it. Tracery is a generative-text library and language implemented in Javascript meant to enable casual users (novice coders and those who do not consider themselves to be coders) to write simple JSON files that encode grammar rules, which then produce complex recursively-expanded text. Tracery has been a very important open source software tool aimed at supporting creative endeavors of many artists and poets. After the initial version was released in 2014, a British artist created the website CheapBotsDoneQuick that hosted bots written in Tracery. CheapBotsDoneQuick in turn created an artbot boom, with more than ten thousand bots currently hosted.
Dr. Compton moved on to create Chancery – a Tracery-style FSM syntax and JS simulation library for multi-state conversational agents – as part of an internship at Google. Chancery was open sourced in October 2017, and Chancery and Tracery became the basis of Dr. Compton’s dissertation, which focused on building AI tools to support “casual creators.” Dr. Compton’s incubator project at CROSS focused on further developing Chancery and developing Tracery 2, the next expansion of Tracery.
The Tracery and Chancery community grew steadily over the last year, despite the challenges of Covid-19. New Tracery bots continued to be made (see https://twitter.com/bmp_hax, https://twitter.com/thecabinetsofb1, https://twitter.com/FineArtThesis) and Tracery was used in numerous games and story-telling experiments. Chancery also saw increased use over the last year. One notable creative use was by the European ambient band Monoganon who used Chancery to build a working conversational agent with Chancery-puppeted animations and dialogue, to accompany a LARP and album release.
As part of Dr. Compton’s wrap up of her CROSS incubator effort, she launched artbot.club, a site for users to host and edit bots in the Summer 2020. Dr. Compton provided a demo of artbot.club as part of the CROSS 2020 Research Symposium (video of session.) The site was also used in a live demonstration at Foundations of Digital Games 2020, where participants used artbot.club‘s Google Sheets’ import-feature to collaboratively build a bot about the conference.
In September 2020, Dr. Compton started a new position as Assistant Professor of Instruction at Northwestern University. In her new role, she intends to continue developing the artbot.club platform, and Tracery and Chancery, and will teach a class using it in Winter 2021. CROSS congratulates Dr. Compton on the successful graduation of her incubator and wishes her well at Northwestern.
About the CROSS Incubator Program
CROSS enables students to follow a career path similar to Sage Weil‘s who created the Ceph storage system as part of his Ph.D. work at UC Santa Cruz. CROSS transfers cutting-edge technology resulting from student research to industry via successful open source projects. CROSS incubator project typically gets its start as PhD or other equivalent project where significant software infrastructure has been developed. Recent PhD graduates (including those from other universities) can propose incubator projects that start out with significant software infrastructure and show strong evidence that the open source software project would meet with great interest by at least one well-established open source software community. For more information on CROSS incubators see our project description page.
This block group hides your featured image, remove this block group to show your featured image again.
