In lieu of an abstract, here is a brief excerpt of the content:

42 Chapter 3 AI Literacy If I had an hour to solve a prob­ lem and my life depended on it, I would use the first 55 minutes determining the proper question to ask. Albert Einstein, Nobel Laureate in Physics (1921) Even if we ­ don’t think we teach digital literacy, we do. Kinda. We teach critical thinking and then hope our students ­ will be as thoughtful and skeptical at ­ doing web searches as we are. We mostly teach critical thinking as some form of disciplinary thinking. ­ There are distinctive tools and skills that distinguish digital literacy and critical thinking in history from computer science, but we want all gradu­ ates to possess shared general thinking skills, like being able to evaluate the credibility and logic of a source or being able to compare conflicting research findings. Thinking with AI ­will be similar: we need both a general skeptical evaluation of every­thing we receive from AI and some understanding of the disciplinary implications. New AI degrees are addressing this combination of the general and specific. For example, the two new AI majors at Purdue Bowen_AI_int_5pgs.indd 42 Bowen_AI_int_5pgs.indd 42 16/02/24 4:27 PM 16/02/24 4:27 PM AI Literacy 43 are offered in the computer science department and claim to blend critical thinking (a handful of philosophy classes) with disciplinary technical skills (Purdue, 2023). Exactly. Just because your ­ Uncle Drew knows the names of ­ every Civil War­ battle in his county, that ­ doesn’t make him a historian. AI also looks likely to amplify existing digital inequities: Male, non–­first generation, White or Asian students are already more likely to be familiar with or have used AI (Goebel, 2023). Faculty ­will need to understand how to use AI in our disciplines and how to teach students to use AI well. Teaching critical thinking mostly within our discipline and hoping that students­ will connect the dots ­ hasn’t always worked. The need for universal AI literacy is also an opportunity to integrate better skepticism , logic, and critical thinking across the curriculum. Faculty have now been charged with determining what skills are essential skills for adapting to a new era of ­ human thinking . While ­ there ­ will be disciplinary variations, AI literacy ­ will need to include how to find the right AI for the job, how to craft better prompts, the ­ causes and danger of hallucinations, and especially the value of iteration. Prob­ lem Formation, Better Questions, and the Liberal Arts AI ­ can’t read your mind. When you start typing in Google, it tries to predict the rest of your search based upon other ­ popular searches. Sometimes this is helpful, but often it is not. With both web searches and AI prompts, better questions yield better results. As the public panic around ChatGPT began, predictions for highly paid ($300,000 per year) ChatGPT “assistants” or “prompt Bowen_AI_int_5pgs.indd 43 Bowen_AI_int_5pgs.indd 43 16/02/24 4:27 PM 16/02/24 4:27 PM [148.135.83.86] Project MUSE (2025-02-16 15:20 GMT) Thinking with AI 44 engineers” started to appear (Their, 2023). Just as previous new technologies created new types of jobs (like Director of Social Media), AI is ­ going to create new types of jobs. The World Economic Forum listed “prompt engineering” as its number-­ one “job of the ­future” in 2023, just ahead of remote truck operator and wind turbine technician (Whiting, 2023). While ­ there ­ will surely be calls for new majors and shifts in computer science curriculum, we should recall that ­ every job is now ­ going to require some AI skills, and the core skill ­ here is about asking better questions, something that has long been central to a liberal arts education. Parents and the public increasingly see college as a place where students acquire the skills needed for a specific profession (despite the finding that youn­ger generations are even more likely to change jobs and professions than ever before). Computer science, therefore, seems like a more practical degree than classics. And indeed, it is easier to see how at least some of the content in a computer science or engineering degree ­ will be immediately valuable—­ that is, in a gradu­ ate’s first job. Now we discover that AI can write computer code, and programmer jobs seem at risk. Students who ­ were especially good at mastering course materials and getting high grades (from any major), but less good at adapting to new prob­ lems and situations­were...

Share