A school tests an AI that automatically grades homework.

The goal is to save teachers time and encourage better learning.

The AI analyzes not only the final answer

but also the reasoning and wording used by the students.

After a few weeks the teachers notice something strange:

Some students constantly receive very high grades.

Yet their solutions are often not very creative.

Meanwhile, students with unusual ideas often receive worse scores,

even when their answers are sometimes better.

Question:
Why can an AI that grades homework unintentionally disadvantage creative solutions?

Solution will be unlocked tomorrow.