The institution has to grade the thinking.
Universities and schools are rewriting their AI policies semester by semester. The harder question sits underneath the policy: what are we teaching students to bring before they ask a tool to bring it for them?
“Schools that keep grading answers will find AI grading them instead. The institutions that grade thinking, reflection and voice will produce the graduates employers compete for, because those graduates will know what they bring before they ask a tool to help them bring it.”
Glenn Llopis’s projection from Conviction Over Code, Volume 2 of The Conviction Papers (2026). A projection, not a research finding.
Each pair below is a shift in what an institution looks for. The left side can now be produced by a tool. The right side can only come from the student.
The same shift applies to faculty and staff. An institution that wants students with a voice has to develop it in its own people first.
Essays, problem sets and exams were designed to show what a student knows. AI has made that evidence easy to produce, and it has exposed how much of the system was built to reward the expected answer.
Universities have been part of our work for two decades, as clients and as hosts of our executive summits.
Students notice whether the adults around them are allowed to think for themselves. That is where we begin.
Alongside Fortune 500 companies and health systems, universities have been part of GLLG’s work for two decades, including as hosts of our executive summits. The same methodology has produced measured results elsewhere:
Full results are on the Proof page.
What is your institution grading for?
Tell us where you are seeing the gap, in students, faculty or leadership, and we will start there.
Developing individual leaders rather than a whole organization? Cohorts and coaching are offered through Conviction Institute.
Other industries