Some Big AI Problems: The Eliza Effect and More
Yes, everyone is talking about AI. However, how do the concerns about AI apply to our classrooms today? Tom Mullaney talks about concerns with: The Eliza effect—where people attribute human characteristics such as trust and credibility to text-generating computers—can be dangerous when combined with the biases and inaccuracies inherent in large language models. It is vital for educators to understand this as we talk about AI with students. There are concerns about using AI as "guest speakers" even for something seemingly "harmless" like "the water cycle." Concerns with humanizing AI. Discussing the "On the Dangers of Stochastic Parrots" paper by Dr. Emily M. Bender et all which discusses the ethical issues and harms of large language models, including bias and environmental racism. Debunking the myth that AI will have values and beliefs. Practical applications of AI in the classroom The challenges of citing generative AI in the classroom.