Marc Watkins writes a thoughtful SubStack about AI in education. It’s primarily focused on higher education, but a lot of it can be applied to K-12.

One of his recent posts, What We Give Up When We Let AI Decide, really hit it out of the ballpark.

It comments on a lot of important issues, particularly the dangers of using AI for student assessment.

I have long shared similar concerns, and feel that AI’s only role in assessment should be in low-stakes online learning games like Wayground or Kahoot, and in low-stakes practice, like Quill for grammar.

I think it’s definitely worth your time to read his piece.