Self-Assessment Questions

AAIR self-assessment questions support the content in this manual and provide an understanding of the type and structure of questions that typically appear on the exam. Often a question will require the candidate to choose the MOST likely or BEST answer among the options provided. Please note that these questions are not actual or retired exam items. Please see About This Manual for more guidance regarding practice questions.

1. Generative adversarial networks (GANs) enhance their ability to create realistic data by:
   1. simultaneous training of components to improve authenticity.
   2. adding more layers to the neural architecture.
   3. employing grouping methods to classify data.
   4. leveraging trial-based learning for iterative rewards.
2. How can organizations BEST ensure that AI solutions align with ethical principles and societal values?
   1. Follow commonly adopted industry standards.
   2. Perform proactive evaluations of ethical impacts.
   3. Schedule regular audits of AI system outcomes.
   4. Ensure transparency of AI models.
3. Which of the following would be the MOST critical consideration when deciding between in-house or cloud infrastructure for an AI application?
   1. Determining the ease of backing out of the agreement with a vendor
   2. Ensuring provider compatibility with business objectives
   3. Evaluating current algorithms used by in-house and vendor developers
   4. Balancing oversight and personalization with adaptability and expenses

Chapter 1 Answer Key — Self-Assessment Questions

1. 1. Training the components simultaneously improves the authenticity of the data generated by generative adversarial networks (GANs).

   2. Increasing layers is related to deep learning (DL) but not specifically to the improvement of GANs.

   3. Grouping methods do not contribute to the adversarial process of GANs.

   4. Trial-based learning is not a method used for GAN data creation.

2. 1. Industry standards do not necessarily take ethical considerations into account and would not ensure alignment with ethical and societal values.

   2. Proactive ethical impact assessments (EIAs), such a fundamental rights impact assessment (FRIA), helps to identify any concerns related to the use of AI solutions and their impact on humans and society. The results of these can help the enterprise ensure alignment with ethical principles and societal values.

   3. Regular audits of outcomes can provide assurance that decisions are being made without bias, but they would not be the best tool to ensure ethical considerations are being adhered to.

   4. While transparency in AI decision making is a key component of ethical use, it alone does not ensure that ethical and societal values are being upheld.

3. 1. While ensuring it will be simple for the enterprise to back out of an agreement with a vendor is key, it is not a primary infrastructure consideration.

   2. Provider compatibility is important for management but secondary in infrastructure decisions.

   3. While it is important to understand the algorithms used by in-house and vendor developers, this is not a critical infrastructure decision.

   4. The principal factor is balancing oversight and personalization with adaptability and expenses. In-house infrastructure offers oversight but may be expensive, while cloud infrastructure offers adaptability and can be cost-efficient, although it involves provider considerations.