Sample essay. Illustrative placeholder text that shows how an essay renders on this site, not the author’s view.
The question
A recommendation engine, a credit model, a résumé screener: each of them makes a claim about a person before the person has done anything. The claim is a probability, and probabilities are not accusations. But the decisions that follow from them are real, and the person on the receiving end usually cannot see the reasoning.
So the question is not whether prediction is allowed. It plainly is; we do it to each other all day. The question is what the predictor owes the predicted.
Three candidate answers
The first answer is accuracy: a model owes nothing but being right often enough. This is the engineer’s answer, and it is not a bad one. A weather forecast does not explain itself and we do not feel wronged by it.
The second answer is an explanation: if a decision affects me, I am entitled to understand why. This is the lawyer’s answer. It sounds obviously correct until you try to write down what counts as an explanation of a model with a billion parameters. “Your score was low” is true and useless. “Feature 4,812 had a large negative weight” is precise and worse.
The third answer is a way out: I am owed the ability to contest the outcome, to supply what the model could not see, to be treated as someone whose story is not finished. This is the answer I find myself drawn to, because it does not require the model to be legible. It requires the institution around the model to be.
Where this leaves a price
A market price is also a prediction about people, aggregated and anonymous. Nobody owes anyone an explanation for it, and that is precisely why it works. The difference is that a price does not decide anything about me; it only tells me what others believe. The models in the first paragraph do decide, and that is where the debt begins.
Sample essay. Placeholder content, not investment or ethical advice.