What generic AI gets wrong with this prompt
This is the most imitated essay in MBA admissions, which means generic AI has seen more mediocre versions of it than any other prompt. That training shows.
It hands you a safe abstraction as your answer
Ask generic AI to help and the what arrives pre-shrunk: impact, excellence, family, helping others. The model averages across every essay it has read, and the average answer is an abstraction nobody can argue with — or feel. A what that could headline ten thousand essays leaves the committee with nothing to remember you by.
Ask your AI — “If my opening statement were posted with no name on it, what kind of person would a stranger picture? How much does that picture match me specifically?”
It answers the what and pads the why
The second clause is the essay, and it’s the clause generic AI physically cannot answer — your actual why lives in memories you never typed into the chat. So the model fills those paragraphs with restatement: what mattering, mattering deeply, mattering in new words. If the why stops at the first comfortable explanation, the essay stops there too.
Ask your AI — “Trace the why in my draft back as far as it goes. Does it end at a specific lived moment, or at a sentence that re-describes the what?”
It drifts into a career narrative
Left alone, generic AI turns this essay into achievement storytelling — the promotion, the project, the wins — because professional narratives dominate both its training and your inputs. But the committee reads this essay with your resume already in hand. 650 words re-proving your career is 650 words not spent on the one question no other document answers.
Ask your AI — “How many sentences of this draft would fit comfortably in a cover letter? What do the remaining sentences reveal that my resume can’t?”
It invents vividness you’ll have to defend
Generic AI decorates origin stories — the rain that fell, the trembling hands, the silence in the room — with sensory details you never provided, because vivid scenes score well in its training data. Invented texture reads as manufactured on the page, and worse, it puts details in your application you can’t speak to in an interview.
Ask your AI — “List every sensory or emotional detail in this draft. Which ones did I actually give you, and which did you add?”