Auto-GPT is a product of an experimental project that makes the use of GPT-4 autonomous. Notably, Auto-GPT emerged and spread rapidly, while the echo of OpenAI's ChatGPT continues. However, there are insufficient studies on this new application in related literature. The purpose of this exploratory case study was to explore the different use cases and experiences of Auto-GPT users. For this purpose, 16 users with an Auto-GPT experience on the GitHub platform were interviewed. Thematic content analysis was performed on the qualitative data. AutoGPT experiences of users can be characterized by learning programs, autonomous applications, conducting research, and writing reports. The results of this study showed that content creation is the most important purpose of using Auto-GPT. As independent research functions of Auto-GPT, users also emphasize data summarization and information organization. However, the participants also pointed out the token limit (inefficiency), forgetting generated tools, and iteration as some prominent limitations of Auto-GPT. It is possible to say that Auto-GPT has a high potential to use in also in educational purpose, but it is still in the development stage.
Auto-GPT is a product of an experimental project that makes the use of GPT-4 autonomous. Notably, Auto-GPT emerged and spread rapidly, while the echo of OpenAI's ChatGPT continues. However, there are insufficient studies on this new application in related literature. The purpose of this exploratory case study was to explore the different use cases and experiences of Auto-GPT users. For this purpose, 16 users with an Auto-GPT experience on the GitHub platform were interviewed. Thematic content analysis was performed on the qualitative data. AutoGPT experiences of users can be characterized by learning programs, autonomous applications, conducting research, and writing reports. The results of this study showed that content creation is the most important purpose of using Auto-GPT. As independent research functions of Auto-GPT, users also emphasize data summarization and information organization. However, the participants also pointed out the token limit (inefficiency), forgetting generated tools, and iteration as some prominent limitations of Auto-GPT. It is possible to say that Auto-GPT has a high potential to use in also in educational purpose, but it is still in the development stage.
Primary Language | English |
---|---|
Subjects | Artificial Intelligence |
Journal Section | Research Articles |
Authors | |
Early Pub Date | August 7, 2023 |
Publication Date | March 10, 2024 |
Published in Issue | Year 2023 |