nerdexam
MicrosoftMicrosoft

GH-300 · Question #16

GH-300 Question #16: Real Exam Question with Answer & Explanation

The correct answer is D: The training data might contain biases or inconsistencies.. {"question_number": 7, "question": "Why might a Generative AI (Gen AI) tool create inaccurate outputs?", "correct_answers": ["D"], "explanation": "Generative AI models learn exclusively from their training data. If that data contains factual errors, outdated information, underrep

Responsible AI

Question

Why might a Generative AI (Gen AI) tool create inaccurate outputs?

Options

  • AThe Gen AI tool is overloaded with too many requests at once.
  • BThe Gen AI tool is experiencing downtime and is not fully recovered.
  • CThe Gen AI tool is programmed with a focus on creativity over factual accuracy.
  • DThe training data might contain biases or inconsistencies.

Explanation

{"question_number": 7, "question": "Why might a Generative AI (Gen AI) tool create inaccurate outputs?", "correct_answers": ["D"], "explanation": "Generative AI models learn exclusively from their training data. If that data contains factual errors, outdated information, underrepresentation of certain domains, or societal biases, the model encodes those flaws and reproduces them in its outputs - a phenomenon often called 'hallucination' or model bias. This is the fundamental, well-documented root cause of Gen AI inaccuracy. Options A and B describe operational/infrastructure issues that are transient and would more likely cause errors or timeouts, not subtly inaccurate outputs. Option C is incorrect because Gen AI tools are not explicitly programmed to prioritize creativity over accuracy; inaccuracy is an emergent artifact of training data quality.", "generated_by": "claude-sonnet", "llm_judge_score": 4}

Topics

#Generative AI#Inaccuracy#Training data#Bias

Community Discussion

No community discussion yet for this question.

Full GH-300 PracticeBrowse All GH-300 Questions