Thanks Jane! Appreciate your writing on these ideas and can't wait to hear about the takeaways from your first-year writing class.
As an aside, I'm working with a colleague at nearby Pomona College to put together some modules that we might use in a first-year writing class at our respective institutions (and other venues). I would be curious to compare notes and brainstorm what information is most useful for that particular set of students too.
Thank you so much for this, Josh. I think another question educators should ask about ai is if AI systems codify the systemic biases and inequities already existing in society, how can we guide students in identifying and consciously counteracting them through technology stewardship?
Asking this question is important because it forces us to confront how AI risks entrenching discrimination observed across existing structures. It challenges educators and students to confront the reality that AI risks amplifying rather than alleviating discriminatory biases established in societal data used to construct models and decision algorithms.
Thanks for the comment, Bechem. Indeed the question of bias is an important and challenging one to address.
I think one place to start is to note, as you did, that these biases are present in the AI models because they exist in the data that they are trained on. It is unlikely that we’ll figure out ways to avoid biases in our algorithms until we figure out how to avoid bias in society. I’m not sure such a thing is possible.
What’s most important is that we articulate what we value and do our best to ensure that our systems support those values. We’ll always have disagreements about values, but it’s important to continue to have the conversations and find as much common ground as we can
Certainly an important area to consider as we use these systems and consider how to improve them.
Thanks for the shout out and for the very useful post!
Thanks Jane! Appreciate your writing on these ideas and can't wait to hear about the takeaways from your first-year writing class.
As an aside, I'm working with a colleague at nearby Pomona College to put together some modules that we might use in a first-year writing class at our respective institutions (and other venues). I would be curious to compare notes and brainstorm what information is most useful for that particular set of students too.
Thank you so much for this, Josh. I think another question educators should ask about ai is if AI systems codify the systemic biases and inequities already existing in society, how can we guide students in identifying and consciously counteracting them through technology stewardship?
Asking this question is important because it forces us to confront how AI risks entrenching discrimination observed across existing structures. It challenges educators and students to confront the reality that AI risks amplifying rather than alleviating discriminatory biases established in societal data used to construct models and decision algorithms.
Thanks for the comment, Bechem. Indeed the question of bias is an important and challenging one to address.
I think one place to start is to note, as you did, that these biases are present in the AI models because they exist in the data that they are trained on. It is unlikely that we’ll figure out ways to avoid biases in our algorithms until we figure out how to avoid bias in society. I’m not sure such a thing is possible.
What’s most important is that we articulate what we value and do our best to ensure that our systems support those values. We’ll always have disagreements about values, but it’s important to continue to have the conversations and find as much common ground as we can
Certainly an important area to consider as we use these systems and consider how to improve them.
Thanks for the shout out. Probably the most thorough post I’ve seen looking at AI and pedagogy.
Thanks Lance!