In a 1,339 teacher experiment, a grade three points too harsh survived review more often when the label said AI. Here is why the order matters in AI grading, and how to mark first on this week’s stack.
A Virginia Tech study recorded how students use ChatGPT for essays. The most common request was the essay itself. Here is what cut it, and how to build that into one assignment this week.
The University of Alaska’s draft AI policy defines human in the loop in one sentence: AI could recommend but a human decides. Here is how to write where that happens into your syllabus, for your students’ AI use and your own.
Dartmouth authorized one AI detector for student work on September 29. Guideline six says a finding needs a constellation of evidence beyond the score. Here is how to build the rest of that case into one assignment before the paper is due.
Andreessen Horowitz launched a school that admits on proof of work and intends to seek approval to charge elite private university tuition in Fall 2028. A Berkeley paper explains why the opening was there. Here is the three line record that puts the evidence back into your own assignments.
A brief unsealed in the news organizations’ case against OpenAI and Microsoft quotes a Microsoft document on a doom loop and Satya Nadella on substitution. Here is how to build the one class session a summary cannot replace.
OpenAI shipped a purpose built legal model with a licensed index over more than 230 million URLs, and published its score: 54.0 percent. The number is not the bad news. Having a number is the good news, and your syllabus does not have one.
The New Mexico Supreme Court held a lawyer in contempt for filing an AI generated brief full of fabricated witnesses. Read the order and the finding is not about the chatbot. It is about the step between making a document and signing it, and your assignment is missing the same step.
Free AI curriculum is arriving on campus with a byline almost nobody reads. The last time this happened it was called learn to code. Here is the four question audit that tells a department whose outcome a unit is actually teaching to.
OpenAI published a claimed Navier Stokes proof and the mathematicians disputing it never once asked for a detector. They asked for a checkable file, a dated record, and a statement of limits. Those three things are assignment design.
AI literacy education built on the SeedStacking™ methodology. Small daily wins that compound into genuine fluency.
We use cookies to improve your experience and analyze site traffic. By continuing, you agree to our Cookie Policy.
The Friday Harvest
One email a week. What changed in AI and higher education, and one thing you can do about it before Monday. Written by a college faculty member who still grades every week.
Get the Friday Harvest