The following text is about technology ethics.
Algorithmic decision-making increasingly affects important life outcomes—from loan approvals to job interviews to criminal sentencing. These systems promise objectivity by removing human bias, but researchers have found that algorithms can encode and amplify existing biases present in their training data. A hiring algorithm trained on historical hiring decisions may learn to discriminate if past hiring was biased. Critics argue that the appearance of mathematical neutrality makes algorithmic bias particularly insidious, harder to detect and challenge than explicit human prejudice.
What concern does the text raise about algorithmic decision-making?
Algorithms cannot process data quickly enough
Algorithms may appear neutral while perpetuating hidden biases
Humans are always more biased than algorithms
Algorithms only affect trivial decisions
Correct Answer: B
Choice B is the correct answer. The text states algorithms "can encode and amplify existing biases" while having "the appearance of mathematical neutrality" that makes bias "particularly insidious, harder to detect."
- Evidence: Bias perpetuation; neutral appearance hides problems.
- Reasoning: The false appearance of objectivity is dangerous.
- Conclusion: Hidden bias behind perceived neutrality is the concern.
Choice A is incorrect because speed isn't discussed. Choice C is incorrect because the text shows algorithms inherit human bias. Choice D is incorrect because "important life outcomes" are affected.