Question needs to be more specific

[Henry C]

14. Question

A technology company is exploring the use of generative AI models, such as large language models (LLMs), for various applications. However, the company is committed to ensuring that these models are deployed in accordance with the principles of responsible AI.

Which of the following are significant challenges of generative AI that need to be considered by the company? (Select THREE)

The option that says: Explainability and transparency is incorrect. While they are relevant in understanding how AI models make decisions, they are not unique challenges specific to generative AI.

The question is asking which of the following are “significant challenges of generative AI”. Since generative AI is a subset of wider AI models, I believe the answer should be correct because the question didnt specify the answer had to be a unique challenge for genAI. Additionally, LLMs are trained with deep neural networks, which are on the lowest scale of explainability of all AI models. I would suggest you change the question to be more specific to generative AI and not Machine learning like “Which of the following are challenges to generative AI specifically, and not a wider range of ML models”.

[Neil-TutorialsDojo]

Hello Henry C,

Good day!

Thank you for your detailed feedback! I understand your point, and I want to clarify that this item is intended to highlight significant challenges specific to generative AI within the context of responsible AI. (https://www.amazon.science/blog/responsible-ai-in-the-generative-era)

While some challenges, like hallucinations and toxicity, are unique to generative AI, others—such as explainability and transparency—are broader principles of responsible AI. Although explainability isn’t exclusive to generative AI, it remains a core dimension of responsible AI that is particularly crucial for generative models.

In conclusion, this item is designed to distinguish between challenges unique to generative AI (like hallucinations and toxicity) and core responsible AI dimensions critical for deploying generative models responsibly. (https://aws.amazon.com/ai/responsible-ai/)

We truly appreciate your insights and will take them into consideration as we refine this item for greater clarity and accuracy.

Thank you again for your valuable feedback!

Best regards,
Neil @ Tutorials Dojo