Qwissit Artificial Intelligence

ISTQB CT-AI

Introduction to AI

60 questions

AI fundamentals

  1. How is the behavior toward new inputs determined in most AI-based systems, particularly ML systems? Easy
  2. Which property do conventional, imperatively programmed systems typically have? Easy
  3. Why are many deep learning models considered a "black box"? Easy
  4. What does the adaptability of self-learning AI-based systems mean for their operation? Easy
  5. Which type of AI covers all AI systems deployed today? Easy
  6. Which term is another name for general AI? Easy
  7. What is the point called at which AI-based systems would transition from general AI to super AI? Easy
  8. Which type of machine learning works with labeled data, for example for prediction and classification? Easy
  9. How does an intelligent agent learn in reinforcement learning? Easy
  10. Which task are convolutional neural networks (CNN) particularly well suited for? Easy
  11. Which AI technology is used for reasoning under uncertainty? Easy
  12. What mainly characterizes generative AI (GenAI)? Easy
  13. Which technology underpins large language models (LLM)? Easy
  14. What are most practical GenAI tools based on today? Easy

Technology and operation

  1. Which type of processor typically outperforms the CPU in ML applications? Easy
  2. What does quantization mean in the context of ML hardware? Easy
  3. What does the abbreviation AIaaS stand for? Easy
  4. What are ML development frameworks used for? Easy

Regulation

  1. What basic approach does the EU AI Act use to regulate AI-based systems? Easy
  2. Which document provides detailed guidance on testing AI-based systems? Easy

AI fundamentals

  1. Which methods do many AI-based systems rely on for problem-solving? Medium
  2. In which domains is the lack of explainability of AI decisions particularly critical? Medium
  3. How can frontier AI be classified? Medium
  4. An AI system recognizes faces with a very high hit rate. Can it also translate texts without further ado? Medium
  5. Which statements about super AI are correct? Medium
  6. Which task do recurrent neural networks (RNN) typically perform? Medium
  7. Which techniques are among the key ML technologies? Medium
  8. What does the term agentic AI describe? Medium
  9. How do generative adversarial networks (GAN) create realistic synthetic data? Medium
  10. Which societal and ethical concerns are there about GenAI? Medium

Technology and operation

  1. Which hardware features typically benefit ML? Medium
  2. What is AI-specific hardware such as ASICs and system-on-a-chip devices best suited for? Medium
  3. What distinguishes neuromorphic processors from conventional processors? Medium
  4. Which advantages do public clouds offer for developing AI models? Medium
  5. What are the pros and cons of hosting smaller models locally on PCs or smartphones? Medium
  6. Which functions do ML development frameworks typically provide? Medium
  7. What is the typical difference between an ML framework with a lower-level API and one with a higher-level API? Medium
  8. Which are examples of hybrid approaches to AI development? Medium

Regulation

  1. Which international soft law instruments provide guidance for responsible AI? Medium
  2. Which requirements does the EU AI Act impose on high-risk systems? Medium

AI fundamentals

  1. A team migrates a rule-based credit check to an ML model. Which consequences for testing follow from the differences between the two kinds of systems? Hard
  2. An authority wants to approve an AI system for medical diagnoses and requires that every decision can be justified. Which difference from conventional systems is the main problem here? Hard
  3. A vendor advertises its new large language model as "general AI" because it writes texts, programs and describes images. How should this be classified? Hard
  4. Which statements about general AI are correct? Hard
  5. An online shop wants to automatically divide customers into groups with similar buying behavior without existing categories. Which type of ML fits? Hard
  6. Which mappings of technology and application area are correct? Hard
  7. An ML system was trained once and has been running unchanged since then. It does not recognize new product categories. How is this to be assessed? Hard
  8. A team needs photorealistic synthetic images as test data. Which generative technologies are suitable for such content? Hard
  9. An agency wants to introduce GenAI for marketing texts. Which statements about opportunities and risks are correct? Hard

Technology and operation

  1. A speech recognition model is to run on low-end smartphones. Which statement is correct? Hard
  2. Which advantages does quantization (computing with fewer bits) have? Hard
  3. A start-up wants to begin a small ML project and hesitates between a computer with a fast CPU and one with a GPU. What is recommended? Hard
  4. A hospital wants to run an AI model with patient data, has strongly fluctuating workloads and a limited budget. Which considerations are correct? Hard
  5. A company wants to train a large model entirely on-premises. What are the consequences? Hard
  6. What are the pros and cons of third-party AI systems compared with privately developed ones? Hard
  7. A team wants to build a model for an embedded device with little memory and has little ML experience. Which criteria for framework selection are particularly relevant here? Hard
  8. Which statements about the evaluation and deployment functions of ML frameworks are correct? Hard

Regulation

  1. What role do technical standards from ISO and IEEE play alongside laws such as the EU AI Act? Hard
  2. Which statements about AI regulation outside the EU and the further development of rules are correct? Hard
  3. A manufacturer of a high-risk AI system for the EU considers extensive testing optional because "penalties are only flat-rate and low anyway". What is the response? Hard

Certifications & exams

Which exam comes next?

You can already practise for 19 exams, and 21 are up for a vote. See all certifications with their official information, cast your vote or suggest your exam.

See all certifications