Classical AI questions from university exams and ML/AI engineer interviews — short answers you can say out loud, with pointers to the topic that explains each one.
Foundations
What is an intelligent agent? What makes it rational?
Describe the task environment of a self-driving car.
Name the five agent architectures.
Search
Compare BFS, DFS, UCS and iterative deepening.
Why is A* optimal, and what conditions does the heuristic need?
How would you design a heuristic for a new problem?
When would you use local search instead of systematic search?
Games & CSPs
Explain alpha-beta pruning and its best case.
Why did Go need MCTS instead of alpha-beta?
How do MRV, forward checking and AC-3 speed up a CSP solver?
Logic, knowledge & planning
What is the difference between soundness and completeness?
Explain forward vs backward chaining with an example.
What is unification and where is it used?
What is the frame problem?
What is STRIPS / PDDL?
Uncertainty, decisions & learning
Why are Bayesian networks more compact than a full joint distribution?
What is an HMM and what is the Viterbi algorithm?
Define an MDP and the Bellman equation.
What's the difference between an MDP and reinforcement learning?
Explain the Q-learning update.
Big picture
What were the causes of the AI winters, and are we in a bubble now?
How would you explain the difference between AI, ML, DL and generative AI to a non-technical person?
What are the main ethical risks of AI systems?