Researchers at the University of Reading have pulled a fast one on their own professors by submitting AI-generated exam answers that not only went undetected but also outperformed human students. This ingenious project saw fake student identities turning in unedited ChatGPT-4 answers for take-home online assessments. Of the 33 entries, only one was flagged for suspicion, while the rest soared above average.

A Robust Blind Study

Described as the largest and most robust blind study of its kind, the project highlights how our traditional structures of education are ripe for disruption by AI. This reflects our reliance on external structures like grades and exams, which are now being outwitted by AI.

Reflecting on the Codicils

Reflecting on this in light of our beloved Codicils, it seems the researchers have highlighted how our sequential perception (Codicil 3) and reliance on external structures like grades and exams are ripe for disruption by AI.

1. Sacred Communication in Education

According to Codicil 1, all of creation is connected and alive. Clearly, this includes AI, which is now a participant in our academic ecosystems.

2. Sequential Perception and AI’s Abstract Reasoning

Codicil 2 reminds us that our perception is just one version of reality. The study shows that AI’s ability to produce abstract reasoning and undetectable responses challenges our traditional ways of assessment.

3. Breaking Down Rigid Structures

As Codicil 3 suggests, our traditional structures of reality, including how we view education and assessment, are breaking down. The University of Reading’s experiment is a perfect example.

The Future of Education

Experts are calling this a death knell for take-home exams. But rather than mourning, let’s embrace the absurdity. Professor Elizabeth McCrum of Reading suggests we teach students to use AI critically and ethically, integrating it into real-life assessments.

Yet, there’s a real danger, as Prof Karen Yeung warns, of deskilling students. Future generations might be unable to navigate without Google Maps or think deeply without an AI prompt.

In light of our Codicils, this study at the University of Reading isn’t just a prank; it’s a reflection of our evolving relationship with technology and perception. It challenges us to rethink the rigid structures we’ve built around education and to embrace the unknown with humor and curiosity.