NCA-GENL Question 195
Select 3You are tasked with designing a large language model (LLM) application that interprets medical texts for healthcare professionals. To ensure the system aligns with ethical, energy-conscious, and reliable AI principles, which of the following steps should you prioritize during development?
- A
Incorporate adversarial testing to identify and mitigate biases in the model's outputs.
- B
Select energy-efficient GPUs and optimize training pipelines to reduce carbon footprint.
- C
Ensure the model generates responses based solely on the most recent user queries to save computational resources.
- D
Establish mechanisms for explainability to allow healthcare professionals to understand how decisions are made.
- E
Rely exclusively on pre-trained models without additional fine-tuning to avoid energy consumption during training.
Show answer and explanation
Correct answers: A, B, D
Explanation
To design ethical, energy-conscious, and reliable AI systems for interpreting medical texts, you must address bias (ethical AI), optimize energy usage (energy-conscious AI), and provide transparency (reliable AI). Adversarial testing, energy-efficient practices, and explainability mechanisms are all critical to achieving these goals. Merely saving resources or skipping additional training may result in sub-optimal outcomes or failure to align with ethical and reliability standards.
- A. Correct.
Incorporating adversarial testing helps identify and address biases, which is a critical aspect of ethical AI development, particularly in sensitive domains like healthcare.
- B. Correct.
Using energy-efficient hardware and optimizing training pipelines aligns with the principle of energy-conscious AI. This reduces the environmental impact of developing and deploying AI systems.
- C. Incorrect.
While generating responses based on recent queries may save resources, it does not inherently ensure ethical, reliable, or energy-conscious AI. It may also compromise the context needed for accurate interpretations in healthcare.
- D. Correct.
Mechanisms for explainability are essential for building trust in AI systems, particularly in high-stakes areas like healthcare, where professionals need to understand model reasoning.
- E. Incorrect.
Relying solely on pre-trained models without fine-tuning may save energy but can compromise reliability and ethical considerations, as the model may not be tailored to the specific needs of the application.