Similarities and Differences Between Human and Artificial Intelligence AIE Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A healthcare startup is developing an AI system to assist doctors in diagnosing rare diseases. The system has been trained on a large dataset of electronic health records. However, the startup has only a small amount of data for a newly discovered disease. The team is debating how to enable the system to recognize this new disease. One option is to train a new model from scratch using only the small dataset. Another option is to use transfer learning from the existing model. The team also considers asking doctors to provide heuristic rules. Which approach is most effective given the constraint of limited data?

   • AUse transfer learning to adapt the existing model to the new disease, leveraging knowledge from the large dataset.
   • BUse the existing model unchanged, assuming it will generalize to the new disease.
   • CRely on doctors to manually encode diagnostic rules for the new disease and bypass the AI model.
   • DTrain a new model from scratch using only the small dataset to avoid bias from the old data.
   Reveal Solution Discussion Select an answer first
2. 2expert · hard

   A medical device company is developing an AI system to detect skin cancer from images. They have a large dataset of labeled images from dermatology clinics. However, the system will be deployed in primary care settings where image quality and lighting conditions differ. The team is debating whether to collect new data from primary care settings or use transfer learning. What is the most important consideration in this decision?

   • ANew data collection is always necessary because AI cannot adapt to new environments.
   • BThe system will generalize well because skin cancer looks the same in all settings.
   • CThe system must be adapted to the new data distribution, which may require new data or transfer learning to maintain accuracy.
   • DTransfer learning is always the best approach because it requires less data.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A bank uses an AI system to approve loan applications. The system analyzes thousands of data points to predict default risk. A customer complains that the system rejected their application without explanation. The bank manager must decide how to respond. What is the most appropriate action?

   • AExplain that the AI uses complex pattern recognition and provide the customer with the key factors that influenced the decision.
   • BBlame the AI system and apologize for its limitations.
   • CManually review the application to provide a more detailed explanation.
   • DTell the customer that the AI's decision is final and cannot be questioned.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which statement accurately describes AI learning compared to human learning?

   • AAI learning is identical to human learning in every aspect
   • BAI learning is typically data-driven and requires large amounts of labeled data
   • CAI learning is experiential and requires physical interaction with the environment
   • DAI learning is based on intuition and common sense
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A research lab is developing an AI system that can pass the Turing test by holding natural conversations. Some researchers claim the system is conscious because it can discuss its own 'feelings' and preferences. The lab's ethicist argues that the system is not conscious. Which evidence best supports the ethicist's position?

   • AThe system's responses are too logical and lack the randomness of human thought.
   • BThe system cannot pass a more advanced version of the Turing test.
   • CThe system does not have a physical body, so it cannot be conscious.
   • DThe system's statements about feelings are generated from patterns in training data, not from subjective experience.
   Reveal Solution Discussion Select an answer first