DirectGPT: A Direct Manipulation Interface to Interact with Large Language Models - Archive ouverte HAL Access content directly
Conference Papers Year : 2024

DirectGPT: A Direct Manipulation Interface to Interact with Large Language Models

Abstract

We characterize and demonstrate how the principles of direct ma- nipulation can improve interaction with large language models. This includes: continuous representation of generated objects of interest; reuse of prompt syntax in a toolbar of commands; manipulable outputs to compose or control the effect of prompts; and undo mechanisms. This idea is exemplified in DirectGPT, a user interface layer on top of ChatGPT that works by transforming direct manipulation actions to engineered prompts. A study shows participants were 50% faster and relied on 50% fewer and 72% shorter prompts to edit text, code, and vector images compared to baseline ChatGPT. Our work contributes a validated approach to integrate LLMs into traditional software using direct manipulation. Data, code, and demo available at https://osf.io/3wt6s.
Fichier principal
Vignette du fichier
DirectGPT__CHI2024_.pdf (1019.44 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04568776 , version 1 (05-05-2024)

Licence

Identifiers

Cite

Damien Masson, Sylvain Malacria, Géry Casiez, Daniel Vogel. DirectGPT: A Direct Manipulation Interface to Interact with Large Language Models. CHI 2024 - ACM Conference on Human Factors in Computing Systems (CHI 2024), ACM, May 2024, Honolulu, United States. ⟨10.1145/3613904.3642462⟩. ⟨hal-04568776⟩
28 View
15 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More