Shaping Human-AI Collaboration: Varied Scaffolding Levels in Co-writing with Language Models
dc.contributor.author | Dhillon, Paramveer | |
dc.contributor.author | Molaei, Somayeh | |
dc.contributor.author | Li, Jiaqi | |
dc.contributor.author | Golub, Maximilian | |
dc.contributor.author | Zheng, Shaochun | |
dc.contributor.author | Robert, Lionel + "Jr" | |
dc.date.accessioned | 2024-02-22T10:35:04Z | |
dc.date.available | 2024-02-22T10:35:04Z | |
dc.date.issued | 2024-02-22 | |
dc.identifier.citation | Paramveer S. Dhillon, Somayeh Molaei, Jiaqi Li, Maximilian Golub, Shaochun Zheng, and Lionel P. Robert. 2024. Shaping Human-AI Collaboration: Varied Scaffolding Levels in Co-writing with Language Models. In Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI ’24), May 11–16, 2024, Honolulu, HI, USA. ACM, New York, NY, USA. | en_US |
dc.identifier.uri | https://doi.org/10.1145/3613904.3642134 | |
dc.identifier.uri | https://hdl.handle.net/2027.42/192479 | en |
dc.description.abstract | Advances in language modeling have paved the way for novel human-AI co-writing experiences. This paper explores how varying levels of scaffolding from large language models (LLMs) shape the co-writing process. Employing a within-subjects field experiment with a Latin square design, we asked participants (N=131) to respond to argumentative writing prompts under three randomly sequenced conditions: no AI assistance (control), next-sentence suggestions (low scaffolding), and next-paragraph suggestions (high scaffolding). Our findings reveal a U-shaped impact of scaffolding on writing quality and productivity (words/time). While low scaffolding did not significantly improve writing quality or productivity, high scaffolding led to significant improvements, especially benefiting non-regular writers and less tech-savvy users. No significant cognitive burden was observed while using the scaffolded writing tools, but a moderate decrease in text ownership and satisfaction was noted. Our results have broad implications for the design of AI-powered writing tools, including the need for personalized scaffolding mechanisms. | en_US |
dc.language.iso | en_US | en_US |
dc.publisher | CHI 2024 | en_US |
dc.subject | Generative AI | en_US |
dc.subject | co-writing | en_US |
dc.subject | Human-AI collaboration | en_US |
dc.subject | writing assistants | en_US |
dc.subject | ChatGPT | en_US |
dc.subject | AI co-writing | en_US |
dc.subject | large language models | en_US |
dc.subject | human-centered interactions | en_US |
dc.subject | AI-powered writing assistance | en_US |
dc.subject | AI scaffolding | en_US |
dc.subject | AI co-writing | en_US |
dc.subject | AI-based writing tools | en_US |
dc.subject | Human-AI collaboration | en_US |
dc.subject | delegation in Human-AI collaboration | en_US |
dc.subject | Chat GPT | en_US |
dc.title | Shaping Human-AI Collaboration: Varied Scaffolding Levels in Co-writing with Language Models | en_US |
dc.type | Conference Paper | en_US |
dc.subject.hlbsecondlevel | Information Science | |
dc.subject.hlbtoplevel | Social Sciences | |
dc.description.peerreviewed | Peer Reviewed | en_US |
dc.contributor.affiliationum | Information, School of | en_US |
dc.contributor.affiliationum | Robotics Department | en_US |
dc.contributor.affiliationum | College of Engineering | en_US |
dc.contributor.affiliationother | UC San Diego | en_US |
dc.contributor.affiliationumcampus | Ann Arbor | en_US |
dc.description.bitstreamurl | http://deepblue.lib.umich.edu/bitstream/2027.42/192479/1/Dhillon et al. 2024 online.pdf | |
dc.identifier.doi | 10.1145/3613904.3642134 | |
dc.identifier.doi | https://dx.doi.org/10.7302/22385 | |
dc.identifier.source | Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI ’24) | en_US |
dc.description.filedescription | Description of Dhillon et al. 2024 online.pdf : Final Version | |
dc.description.depositor | SELF | en_US |
dc.working.doi | 10.7302/22385 | en_US |
dc.owningcollname | Information, School of (SI) |
Files in this item
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.