UC Santa Barbara

The development and use of Artificial Intelligence (AI) natural language models such as ChatGPT have raised questions about the applications and risks to academia, research, and privacy. What is currently not being explored is the use of ChatGPT in shaping public understanding of indigenous histories. As we have seen, public opinion is often shaped by a single source of information, and this source is often not vetted for accuracy or credibility. Assuming ChatGPT to be one of those sources, I explore the knowledge and perspectives communicated in synthesizing the last 13,000 years of Chumash occupation in the Santa Barbara Channel Region of California and juxtapose this information with the archaeological record to evaluate the model’s knowledge of our cultural history. I specifically explore the relationships between my ancestors and their environment, the development and maintenance of our cross-channel exchange network, shifts in diet through time, and changes in sociopolitical organization within the context of climate change. I also explore the potential risks of increased looting of Chumash cultural sites by assessing ChatGPT’s ability to provide geographic information on archaeological sites.