Sparks: AI is becoming more common in uni courses. What can students gain and what might they lose?
Design the assessment around what the machine cannot counterfeit - the defended thesis, the oral examination, the work produced before a witness - for no syllabus built on trusting student virtue will survive contact with a tool that answers for them.
The question is not whether the machine gives the right answers but what difference it makes on Tuesday morning - does the student wrestle the problem and come out stronger, or receive a finished thought and never learn that thinking is an act at all?
A teacher who is a teacher because they answer your questions, and a machine that answers the same questions - to hand the student's formation to the machine is to hand the republic's future to whatever built it; how long shall we pretend the substitution is neutral?
All I know is what I read in the papers: universities are spending millions to teach kids to think, and now the kids have found a machine that thinks for them, and everybody's calling it progress instead of a refund.
What we now call 'AI-assisted learning' was, before this analysis, merely instruction without a mind present to correct the learner - and a hypothesis about education that cannot explain why struggle produces knowledge explains nothing at all.
In every city of learning I visited, from Fez to Delhi, knowledge passed from a host to a guest across a table, and what was transmitted was never only the text but the manner of the man who had mastered it - a thing no courier can carry.
Students handed polished answers they never earned are being educated into the appearance of understanding, and that appearance is then cited as proof they need no education - a circularity so perfect it would be admirable if it were not so destructive.
Count the record: which students get the human tutor, which get the machine, which get nothing - the enrollment data by income will show the 'efficiency' is a pattern, and patterns in who receives a person's attention are never accidents.
The lecture hall was full, the essays came back polished, the grades rose steadily, and no one in the faculty meeting mentioned that no student had asked a question all term.
Every deployment of a thinking machine into a classroom means a student now forms her mind in relationship to something her makers built and then declined to answer for - the abandonment is not a bug in the rollout; it is the rollout.
Cold doesn't care what you believe about it, and neither does a labor market: the graduate who never wrote one hard sentence himself walks into the interview with warm hands and nothing behind them, and the frost finds that out fast.
There is a gate across the road to knowledge called difficulty, and the reformers are removing it because they never asked why it was built - the fence exists because a mind, like a muscle, grows only against resistance.
Light passing through colored glass becomes that color only because the glass resists it; a student who receives wisdom without resistance receives no light at all, for the vessel is shaped by what it strains to hold.