ThatQuiz Test Library Take this test now
Artificial Intelligence - Exam
Contributed by: Squires
  • 1. What does AI stand for?
A) Automated Intelligence
B) Advanced Intelligence
C) Analogous Integration
D) Artificial Intelligence
  • 2. What is the Turing Test?
A) A test to measure a machine's processing speed
B) A test to evaluate a machine's physical strength
C) A test of a machine's ability to exhibit intelligent behavior indistinguishable from a human
D) A test to determine the power consumption of a machine
  • 3. Which programming language is commonly used in AI development?
A) Ruby
B) Java
C) C++
D) Python
  • 4. What is the concept of machine learning?
A) A method to improve network security
B) A technique to manually program machines
C) A subset of AI that enables machines to learn from data
D) A process of assembling hardware components
  • 5. What does RNN stand for in the context of neural networks?
A) Robust Neuron Navigator
B) Rapid Notification Node
C) Recurrent Neural Network
D) Regular Numeric Notation
  • 6. What is the term 'singularity' in the context of AI?
A) A type of machine learning algorithm
B) A measure of data complexity
C) A weather manipulation technique
D) A hypothetical future point at which AI surpasses human intelligence and control
  • 7. What is computer vision in AI?
A) Testing computer hardware components
B) Mimicking human vision and identifying objects in images or videos
C) Generating random pixel patterns
D) Analyzing audio signals
  • 8. What is a chatbot?
A) A program for virtual reality gaming
B) A program for music composition
C) A program that simulates conversation with human users
D) A program for graphic design
  • 9. What is the Dijkstra algorithm used for in AI?
A) Detecting errors in data
B) Optimizing computer memory usage
C) Generating random numbers
D) Finding the shortest path in a graph
  • 10. What does NLP stand for in the context of AI?
A) Nonlinear Linguistic Pattern
B) Neural Learning Protocol
C) Networked Logistic Performance
D) Natural Language Processing
  • 11. In which year was artificial intelligence founded as an academic discipline?
A) 1965
B) 1956
C) 1972
D) 1980
  • 12. Which of the following is NOT a traditional goal of AI research?
A) Learning
B) Knowledge representation
C) Quantum computing
D) Reasoning
  • 13. Which company aims to create artificial general intelligence (AGI)?
A) IBM
B) Intel
C) Microsoft
D) OpenAI
  • 14. What is the name of the architecture that accelerated AI growth further after 2017?
A) Transformer architecture
B) Perceptron
C) Recurrent neural network
D) Convolutional neural network
  • 15. Which high-profile application of AI is used by YouTube, Amazon, and Netflix?
A) Virtual assistants
B) Recommendation systems
C) Advanced web search engines
D) Autonomous vehicles
  • 16. Which field is NOT mentioned as one that AI draws upon?
A) Linguistics
B) Psychology
C) Astronomy
D) Neuroscience
  • 17. Which technique is NOT used by AI researchers to reach their goals?
A) Artificial neural networks
B) Quantum entanglement
C) Formal logic
D) State space search
  • 18. What decade saw an AI boom coinciding with advances in generative AI?
A) 2010s
B) 1990s
C) 2020s
D) 2000s
  • 19. What are some potential long-term effects of AI that have prompted discussions?
A) Reduced software complexity
B) Decreased computational power
C) Existential risks
D) Lower energy consumption
  • 20. What is a major limitation of early AI algorithms in reasoning and problem-solving?
A) They were unable to process any form of incomplete information.
B) Early AI could not handle logical deductions.
C) These algorithms required human intervention for every step.
D) They experience a 'combinatorial explosion' where they become exponentially slower as problems grow.
  • 21. How do humans typically solve most of their problems, as opposed to early AI models?
A) Humans use fast, intuitive judgments rather than step-by-step deduction.
B) Humans rely solely on logical deductions similar to early AI models.
C) Humans use a combination of intuition and probabilistic reasoning exclusively.
D) Humans solve problems by following pre-defined algorithms.
  • 22. In automated planning, what does the agent have?
A) A specific goal.
B) Multiple goals to achieve simultaneously.
C) Randomly assigned tasks with no particular order.
D) No clear objective or preference.
  • 23. Which type of machine learning involves finding patterns in data without guidance?
A) Supervised learning
B) Transfer learning
C) Reinforcement learning
D) Unsupervised learning
  • 24. In supervised learning, what is the main difference between classification and regression?
A) Classification uses neural networks while regression does not.
B) Classification is a type of unsupervised learning.
C) Regression requires more data than classification.
D) Classification predicts categories while regression deduces numeric functions.
  • 25. What modern NLP technique represents words as vectors encoding their meaning?
A) Machine translation
B) Word embedding
C) Speech synthesis
D) Information retrieval
  • 26. Which deep learning architecture uses an attention mechanism?
A) Transformers
B) Generative pre-trained transformers (GPT)
C) Recurrent neural networks (RNNs)
D) Convolutional neural networks (CNNs)
  • 27. What is a moderate success related to affective computing?
A) Speech recognition.
B) Object tracking.
C) Textual sentiment analysis.
D) Image classification.
  • 28. Which type of search is used in game-playing programs like chess?
A) Particle swarm optimization.
B) Gradient descent.
C) Adversarial search.
D) Local search.
  • 29. What technique does local search use to refine an initial guess incrementally?
A) Mathematical optimization.
B) Backpropagation algorithm.
C) Means-ends analysis.
D) Swarm intelligence algorithms.
  • 30. Which variant of local search is commonly used to train neural networks?
A) Particle swarm optimization.
B) Ant colony optimization.
C) Gradient descent.
D) Evolutionary computation.
  • 31. What is the process called when proving a new statement from given true statements?
A) Evolutionary computation.
B) Deductive reasoning.
C) Inductive reasoning.
D) Particle swarm optimization.
  • 32. What is the main challenge with problem-solving in first-order logic?
A) It uses swarm intelligence algorithms.
B) It requires gradient descent for optimization.
C) It assigns degrees of truth between 0 and 1.
D) Inference is undecidable, making it intractable.
  • 33. Which swarm intelligence algorithm is inspired by bird flocking?
A) Gradient descent.
B) Particle swarm optimization.
C) Ant colony optimization.
D) Evolutionary computation.
  • 34. Which of the following tools can be used for reasoning using Bayesian inference?
A) Bayesian networks
B) Kalman filters
C) Markov decision processes
D) Dynamic decision networks
  • 35. What algorithm is commonly associated with learning in Bayesian networks?
A) Decision analysis
B) Mechanism design
C) Expectation–maximization algorithm
D) Information value theory
  • 36. Which method was the most widely used analogical AI until the mid-1990s?
A) K-nearest neighbor algorithm
B) Naive Bayes classifier
C) Support vector machine
D) Decision tree
  • 37. Which classifier is reportedly the most widely used learner at Google due to its scalability?
A) Decision tree
B) Support vector machine
C) Naive Bayes classifier
D) K-nearest neighbor algorithm
  • 38. What type of AI application uses pattern matching to determine the closest match?
A) Neural networks
B) Controllers
C) Bayesian networks
D) Classifiers
  • 39. Which symbolic machine learning algorithm is considered the simplest and most widely used?
A) K-nearest neighbor algorithm
B) Decision tree
C) Support vector machine
D) Naive Bayes classifier
  • 40. Which probabilistic method can be used for filtering, prediction, and smoothing?
A) Decision analysis
B) Dynamic decision networks
C) Hidden Markov models
D) Game theory
  • 41. What type of AI application is exemplified by 'if diamond then pick up'?
A) Neural networks
B) Controllers
C) Classifiers
D) Bayesian networks
  • 42. Which tool from probability theory and economics helps in decision-making and planning?
A) Dynamic Bayesian networks
B) Kalman filters
C) Decision theory
D) Expectation–maximization algorithm
  • 43. Which probabilistic model is used for perception systems to analyze processes over time?
A) Mechanism design
B) Dynamic Bayesian networks
C) Game theory
D) Markov decision processes
  • 44. Which algorithm is most commonly used for training neural networks?
A) Backpropagation
B) Stochastic gradient descent
C) Gradient descent
D) Forward propagation
  • 45. In which direction does the signal pass in feedforward neural networks?
A) Randomly
B) Backwards
C) Both directions
D) Only one direction
  • 46. In image processing, what might lower layers of a deep learning model identify?
A) Faces
B) Edges
C) Whole objects
D) Digits
  • 47. What is the primary function of Generative Pre-trained Transformers (GPT)?
A) Generate text based on semantic relationships between words.
B) Predict future stock market trends.
C) Translate languages in real-time.
D) Analyze and interpret images.
  • 48. Which of the following is NOT a current model or service based on GPT?
A) Gemini
B) Claude
C) Prolog
D) ChatGPT
  • 49. Which software was commonly used with AI-specific enhanced GPUs?
A) PyTorch.
B) Keras.
C) Scikit-learn.
D) TensorFlow.
  • 50. Who is credited with identifying Moore's Law?
A) Jensen Huang.
B) John McCarthy.
C) Gordon Moore.
D) Alan Turing.
  • 51. What trend describes the faster improvements in GPU performance compared to Moore's Law?
A) Moore's law.
B) Huang's law.
C) Bell's law.
D) Gibson's law.
  • 52. Which company developed AlphaFold 2?
A) DeepMind
B) Microsoft
C) IBM
D) Google
  • 53. Which AI system achieved grandmaster level in StarCraft II?
A) Watson
B) AlphaStar
C) MuZero
D) Deep Blue
  • 54. In what year did AlphaFold 2 demonstrate the ability to approximate protein structures quickly?
A) 2021
B) 2019
C) 2024
D) 2023
  • 55. Which AI system was introduced by Google DeepMind in 2024?
A) Pluribus
B) SIMA
C) MuZero
D) AlphaStar
  • 56. Which virtual assistant is developed by Apple?
A) Google Assistant
B) Cortana
C) Siri
D) Alexa
  • 57. What role oversees the deployment of AI in organizations?
A) Chief Technology Officer (CTO)
B) Chief Automation Officer (CAO)
C) Chief Data Officer (CDO)
D) Chief Information Officer (CIO)
  • 58. Which AI system defeated two of the greatest Jeopardy! champions in 2011?
A) AlphaGo
B) MuZero
C) Deep Blue
D) Watson
  • 59. What type of games does Pluribus handle?
A) Chess and Go.
B) Real-time strategy games.
C) Imperfect-information games like poker.
D) Jeopardy! quiz shows.
  • 60. Which AI system was developed to play chess, Go, or Atari games?
A) Deep Blue
B) Watson
C) MuZero
D) AlphaStar
  • 61. Which company developed Qwen2-Math?
A) Alibaba Group
B) Microsoft
C) Google DeepMind
D) OpenAI
  • 62. What was the accuracy of Qwen2-Math on the MATH dataset?
A) 53%
B) 84%
C) 75%
D) 90%
  • 63. Which technique did Microsoft propose in January 2025 to improve mathematical problem-solving?
A) Gemini Deep Think
B) AlphaTensor
C) Qwen2-Math
D) rStar-Math
  • 64. What percentage of AIME 2024 problems did Qwen-7B solve using rStar-Math?
A) 53%
B) 75%
C) 90%
D) 84%
  • 65. Which model achieved gold medal results in the International Math Olympiad of 2025?
A) Qwen-7B
B) rStar-Math
C) AlphaTensor
D) Gemini Deep Think
  • 66. What does topological deep learning integrate?
A) Natural language processing
B) Monte Carlo tree search
C) Various topological approaches
D) Probabilistic models
  • 67. In which month and year did Google introduce an AI Mode?
A) May 2025
B) July 2024
C) February 2023
D) December 2017
  • 68. Which company introduced Copilot Search under the name Bing Chat?
A) Apple
B) Amazon
C) Google
D) Microsoft
  • 69. What percentage of companies reported incorporating AI in some offerings or processes according to a 2017 survey?
A) 20%
B) 50%
C) 10%
D) 5%
  • 70. What technique is used by AI developers to preserve privacy while obtaining data?
A) Cloud storage
B) Differential privacy
C) Data encryption
D) Blockchain technology
  • 71. Which companies dominate the commercial AI scene?
A) Nike, Adidas, Puma, Reebok
B) Tesla, SpaceX, Uber, Lyft
C) Coca-Cola, PepsiCo, Red Bull, Monster
D) Alphabet Inc., Amazon, Apple Inc., Meta Platforms, Microsoft
  • 72. What was the amount spent on authorized AI-generated content during the 2024 Indian elections?
A) $100 million
B) $25 million
C) $10 million
D) $50 million
  • 73. What was the estimated investment into AI infrastructure and data centers in the US by 2030 according to a McKinsey & Company report?
A) $3.5 trillion
B) $1.5 trillion
C) $2.7 trillion
D) $4.0 trillion
  • 74. By what year did the International Energy Agency project that power demand for data centers and AI might double?
A) 2026
B) 2025
C) 2030
D) 2028
  • 75. How much more electrical energy does a ChatGPT search use compared to a Google search?
A) 20 times
B) 10 times
C) 5 times
D) 15 times
  • 76. What percentage of US power was forecasted for data centers to consume by 2030 according to a Goldman Sachs Research Paper?
A) 5%
B) 8%
C) 10%
D) 12%
  • 77. What percentage of US power did data centers consume in 2022?
A) 3%
B) 10%
C) 7%
D) 5%
  • 78. Which nuclear power plant did Microsoft announce an agreement to reopen for its energy needs?
A) Palisades Nuclear reactor
B) Fukushima
C) Susquehanna
D) Three Mile Island
  • 79. Which company's application to supply electricity from the Susquehanna nuclear power station was rejected by FERC?
A) Amazon
B) Microsoft
C) Talen Energy
D) Constellation Energy
  • 80. Which country lifted its ban on the opening of data centers in 2022 due to electric power concerns?
A) United States
B) Japan
C) Taiwan
D) Singapore
  • 81. What is the potential reduction in energy sector emissions estimated for AI?
A) 7%
B) 3%
C) 10%
D) 5%
  • 82. What was the primary goal given to recommender systems on platforms like YouTube and Facebook?
A) Maximizing user engagement
B) Promoting accurate information
C) Enhancing content diversity
D) Reducing misinformation spread
  • 83. What term describes the phenomenon where users are exposed to multiple versions of the same misinformation?
A) Echo chambers
B) Confirmation bias
C) Filter bubbles
D) Information overload
  • 84. Who expressed concern about AI enabling authoritarian leaders to manipulate electorates?
A) Geoffrey Hinton
B) Bill Gates
C) Tim Cook
D) Elon Musk
  • 85. What term is used for malicious use of AI-generated content to create misinformation or propaganda?
A) Synthetic media
B) Deepfakes
C) Faux images
D) AI clones
  • 86. What solution have AI researchers suggested to combat online deception by AI models?
A) Personhood credentials
B) AI ethical guidelines
C) Blockchain verification
D) Digital signatures
  • 87. What was the error rate for both whites and blacks in the COMPAS system?
A) Exactly 61%
B) 50%
C) 80%
D) 75%
  • 88. What percentage of AI engineers are black?
A) 10%
B) 25%
C) 50%
D) About 4%
  • 89. Which category of fairness focuses on the outcomes and seeks to compensate for statistical disparities?
A) Representational fairness
B) Predictive fairness
C) Distributive fairness
D) Procedural fairness
  • 90. What type of weapon is described as a machine that locates, selects, and engages human targets without human supervision?
A) Conventional firearm
B) Lethal autonomous weapon
C) Drones used for surveillance
D) Cybersecurity tool
  • 91. In what year did 30 nations, including China, support a ban on autonomous weapons under the United Nations' Convention on Certain Conventional Weapons?
A) 2013
B) 2014
C) 2016
D) 2015
  • 92. What percentage of U.S. jobs did Michael Osborne and Carl Benedikt Frey estimate are at 'high risk' of potential automation in the 2010s?
A) 47%
B) 9%
C) 60%
D) 25%
  • 93. According to an OECD report, what percentage of U.S. jobs were classified as 'high risk' for automation?
A) 15%
B) 47%
C) 9%
D) 30%
  • 94. What percentage of jobs for Chinese video game illustrators were reported to be eliminated by generative AI in April 2023?
A) 50%
B) 90%
C) 30%
D) 70%
  • 95. Who coined the term 'Friendly AI'?
A) Wendell Wallach
B) Stuart J. Russell
C) Stephen Hawking
D) Eliezer Yudkowsky
  • 96. What is another name for the field of machine ethics?
A) Moral robotics
B) Computational morality
C) Artificial intelligence ethics
D) Ethical computing
  • 97. Who proposed the concept of 'artificial moral agents'?
A) Wendell Wallach
B) Stephen Hawking
C) Eliezer Yudkowsky
D) Stuart J. Russell
  • 98. What is one potential risk of open-weight AI models?
A) Built-in security measures can be trained away until ineffective.
B) They cannot be used for commercial purposes.
C) They require constant internet connectivity.
D) Their architecture and parameters are kept secret.
  • 99. Which program developed by DeepMind defeated the world champion Go player in 2015?
A) AlphaGo
B) DALL-E
C) GPT-3
D) ChatGPT
  • 100. What percentage of newly funded startups claimed to be AI companies in 2024?
A) 75%
B) 50%
C) 5%
D) 22%
Created with That Quiz — a math test site for students of all grade levels.