Crap Detection and AI

When my daughter first started using search engines when she was in middle school, I sat down with her and demonstrated how she could summon millions of answers to any question within a couple seconds, but that now there was no guarantee that any of them would be true or accurate -- that determination was now up to her. I ended up writing my first piece on crap detection in 2009. In 2010, I wrote about it for educators. In 2011, I made a video. Later, I included crap detection in my Stanford course on Social Media Literacies. And my 2012 book, Net Smart, included a chapter on it.

Shortly after my book and my Stanford course, Sam Wineburg, a Stanford historian and education professor, invited me to walk around campus. To my delight and astonishment, he presented me with several thousand dollars for my work that inspired his initiative to teach students how to use Internet search with a critical eye. That work grew into Stanford's Digital Inquiry Group, led by historian Sam Wineburg.

All of which is to preface this recent post by Wineburg about the need for AI crap detection skills for students today:

Education should teach students to grapple with complexity. AI is designed to avoid it. This mismatch is yet another reason to slow roll the rush to put AI in students’ hands. It also points to a question teachers should ask themselves when evaluating how to integrate AI into the classroom: Can I use this tool in a way that models the thinking I want to teach my students?

For more than two decades, the Digital Inquiry Group and its earlier iteration at Stanford University have created curriculum that teaches students to read and think like historians, centered on document-based inquiry. After a 2016 study showed that young people struggle to evaluate online sources, our research group developed curriculum to help students separate fact from fiction on the internet.