AI Killed the Take-Home Essay; COVID Helped
Source: Steven Mintz Substack
Author: Steven Mintz
Original source: https://stevenmintz.substack.com/p/ai-killed-the-take-home-essay-covid
Source type: essay
Private backup: the full article text is archived in the private repository at archives/articles/stevenmintz-substack-com-ai-killed-the-take-home-essay-covid.source.md. It is not published on the public Quartz site.
Summary
Steven Mintz argues that higher education has entered a post-COVID, AI-saturated instructional environment where traditional take-home essays and other unsupervised assessments can no longer be trusted as straightforward evidence of student learning. He links AI to broader pressures around attendance, reading, attention, economic precarity, accommodations, and weakened academic habits. His response is to redesign courses around mandatory attendance, handwritten in-class prompts, student-led discussion, oral presentations, and observable demonstrations of understanding.
Pull quotes
Observable learning
“The work that matters will be the work I witness.”
— Mintz, on shifting assessment toward in-class evidence.
Assessment evidence breaks down
“I could no longer trust the primary evidence on which I had always relied to understand what students were learning.”
— Mintz, explaining why unsupervised essays no longer felt reliable.
The boundary problem
“The phrase assumes a clear boundary between autonomous and assisted thinking, but that boundary has become philosophically unstable.”
— Mintz, on the instability of “own work” in AI-mediated writing.
Performance cannot be outsourced
“But the performance itself cannot be outsourced.”
— Mintz, arguing that live explanation and defense still reveal understanding.
Big ideas
- AI simulations need clear boundaries for learning
- Students need to bring the purpose; AI should not supply it for them
Claims
- Take-home essays are no longer reliable evidence by themselves
- AI-assisted homework requires redesign, not just policing
Key evidence and examples
- Mintz plans to replace take-home essay-centered grading with short handwritten in-class prompts, mandatory attendance, student-led sessions, discussion, presentations, and live questioning.
- He reports seeing polished but generic undergraduate essays that were detached from course specifics and sometimes referenced sources not assigned or discussed.
- The article identifies converging pressures: pandemic habits, fractured attention, economic precarity, accommodations, weakened reading expectations, and AI-assisted writing.
- AI destabilizes the boundary between assistance and outsourcing because outlines, summaries, revisions, and polished prose can all blur authorship.
- Oral defense, facilitation, and live explanation are presented as harder to outsource because real understanding becomes visible.
Education relevance
This is useful for assessment redesign, academic integrity policy, post-COVID pedagogy, writing instruction, and the move toward observable learning in AI-era classrooms.
Durability note
This is a durable assessment-design argument even though the COVID and generative-AI context is historically specific: when outside-work evidence becomes unreliable, schools need more observable demonstrations of thinking.