Major Milestones in AI Evolution AIE Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A company is developing an AI system to predict equipment failures in a factory. The team is considering two approaches: a deep learning model that learns from sensor data, and an expert system that encodes the knowledge of maintenance engineers. The company has a limited budget and needs a system that can be deployed quickly. The team lead notes that the expert system can be built with input from engineers, while the deep learning model requires a large amount of historical failure data, which is scarce. Which approach should the team choose, considering the constraints?

   • AChoose the expert system because it can be built quickly using the engineers' knowledge and does not require large datasets.
   • BChoose the deep learning model because it is the most modern approach.
   • CChoose the deep learning model because it will be more accurate than the expert system.
   • DChoose both approaches and run them in parallel, regardless of the cost.
   Reveal Solution Discussion Select an answer first
2. 2expert · hard

   A company is developing an AI system to detect fraudulent transactions. The team is considering two approaches: a rule-based expert system that encodes known fraud patterns, and a deep learning model that learns from historical transaction data. The company has a large dataset of labeled transactions, but the fraud patterns are constantly evolving. The team lead notes that the expert system can be updated manually, while the deep learning model can be retrained. Which approach is more suitable for adapting to evolving fraud patterns?

   • ABoth approaches are equally suitable, as they can both adapt to new patterns.
   • BThe deep learning model, because it can be retrained on new data to automatically learn new fraud patterns.
   • CThe expert system, because it can be updated by domain experts.
   • DNeither approach is suitable, as fraud detection is impossible with AI.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A healthcare startup is developing a diagnostic tool. They are considering two approaches: a rule-based expert system that encodes physician knowledge, and a deep learning model trained on thousands of medical images. The team needs to decide which approach to use, knowing that the expert system will be easier to audit but harder to maintain, while the deep learning model may find patterns humans miss. Which consideration, rooted in the historical evolution of AI, is most relevant to this decision?

   • AThe choice depends on the trade-off between explainability and the ability to learn complex patterns from data, a tension that has shaped AI since the shift from symbolic to connectionist approaches.
   • BDeep learning models are always more accurate than expert systems, so the team should choose it.
   • CThe team should always prefer the approach that was developed first, as it is more mature.
   • DExpert systems are the only approach that has been successfully applied in healthcare, so the team should choose it.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which of the following sequences correctly orders major milestones in AI evolution from earliest to most recent?

   • ADartmouth workshop → Turing test proposed → Deep learning → Expert systems → Generative models
   • BDeep learning → Turing test proposed → Dartmouth workshop → Expert systems → Generative models
   • CExpert systems → Turing test proposed → Dartmouth workshop → Generative models → Deep learning
   • DTuring test proposed → Dartmouth workshop → Expert systems → Deep learning → Generative models
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A healthcare startup is building a diagnostic support system that must explain its reasoning to clinicians. The team is debating between a rule-based approach and a modern deep learning model. Which historical milestone best supports the rule-based approach's suitability for this requirement?

   • AThe invention of the perceptron in 1958, which demonstrated that machines could learn from labeled examples.
   • BThe introduction of the Turing test in 1950, which defined machine intelligence as the ability to imitate human conversation.
   • CThe rise of deep learning after 2012, which enabled high-accuracy pattern recognition from large datasets.
   • DThe development of expert systems in the 1970s, which encoded human expertise into if-then rules for explainable decision support.
   Reveal Solution Discussion Select an answer first