If you’re teaching fifth through twelfth grade humanities or social sciences using an AI as an underassistant agent, try this out.
Part 1: Ask students to hazard a guess as to how many incumbent Presidents have been re-elected for a second term. Discuss with them why such a statistic might be interesting to know. Ask them how they might go about finding out. Agree with the first student who says “That’s a job for an LLM like GPT4.”
Part 2: Type a prompt into the bot, but first challenge students to write a prompt getting at the desired information: Name in alphabetical order each US incumbent President who was re-elected. Include dates, party affiliation, and a caption or sentence revealing the theme of the first term
Optional but recommended if you’re using a bot that denies commands for political analysis. When I did it, I proactively defused raising the hackles of sentiment and ethical algorithms by describing myself as a centrist with sympathies tending in both directions but pragmatically more liberal. I told the bot my interest stems from an incumbent President now running for President.
I wanted to know two things by the end of this chat. First, what is it about those re-elected Presidents they shared that got them re-elected? Second, how many of them had successful second terms? Are there lessons to be learned from analyzing their records?
Part 3: Challenge students to write a prompt to get the list of re-elected Presidents by name, party affiliation, dates, and caption or slogan. Then enter the command. The output should list 16 Presidents. Assign small groups with a bot to two-four Presidents. Cover all 16. They work together to develop and verify information in regard to the following:
What is it about those re-elected Presidents that got them re-elected?
How many of them had successful second terms?
Are there lessons for voters to be learned from analyzing their records?
What insights into Presidential elections can you share, if any?
What qualities do good first term Presidents show?
Part 4: Provide report out time. Each group should summarize bot output and share websites on the class screen to go more deeply into facts of the President’s terms. Each individual should share orally one of the group’s insights.
Part 5: Assign a reflective memo of one page including a brief description of what was done, insights into using a bot collaboratively, and how verification helped with the report out.
This is rough. I think if I were teaching I’d love to try it after some tweaks if I could manage it without detritus flying out of the sky from the South. In that event I might show a movie. Just kidding.