Professor Emilio Herrera had been dead for three years, yet his final problem set haunted the graduate students of the University of Seville like a ghost story told in the dark.
She flipped to Problem 4.22: "The number of coding errors in a software module follows a Poisson distribution with mean λ. Derive the MLE of λ given a sample of bug reports from five developers."
The file opened not as a PDF, but as a living document. The first page read: "Estimado estudiante: Usted ha encontrado las respuestas. Pero aquí, las preguntas son más importantes. Cada problema resuelto es una semilla. Plántala mal, y obtendrás un error. Plántala bien, y obtendrás una verdad." (Dear student: You have found the answers. But here, the questions are more important. Each solved problem is a seed. Plant it wrong, and you will get an error. Plant it right, and you will get a truth.)
She plugged it in.
Elena froze. The navigation module failure had cost the university's satellite project two months of delays. She had been a junior analyst on that project. Herrera had known she would one day open this file.
Then she made a new file. She labeled it: