Question 1

What is Artificial Intelligence, and what is the difference between Weak AI and Strong AI?

Accepted Answer

Artificial Intelligence is the simulation of human intelligence processes by computer systems, including learning, reasoning, and self-correction. Weak AI (Narrow AI) is designed and trained for a specific task (e.g. chess engines, voice assistants). Strong AI (Artificial General Intelligence / AGI) is a theoretical system that possesses cognitive abilities equal to or exceeding a human, capable of solving any general problem it encounters.

Question 2

Explain the Turing Test.

Accepted Answer

Proposed by Alan Turing in 1950, the Turing Test is a test of a machine's ability to exhibit intelligent behavior indistinguishable from that of a human. A human evaluator engages in a natural language conversation with a human and a machine in separate rooms. If the evaluator cannot reliably tell the machine from the human, the machine passes the test.

Question 3

What is a Heuristic function in AI search algorithms?

Accepted Answer

A Heuristic function $h(n)$ estimates the cost from the current node $n$ to the goal state. It guides search algorithms towards the goal more efficiently than blind searches (like Breadth-First or Depth-First Search) by pruning search branches that appear less promising.

Question 4

Explain the A* Search algorithm.

Accepted Answer

A* Search is a pathfinding and graph traversal algorithm known for its completeness and efficiency. It selects the next node $n$ to expand based on the evaluation function $f(n) = g(n) + h(n)$, where $g(n)$ is the exact cost from the start node to node $n$, and $h(n)$ is the heuristic estimate of the cost from node $n$ to the goal. A* is optimal if the heuristic is admissible (never overestimates the actual cost).

Question 5

Explain BFS vs DFS in state space search.

Accepted Answer

BFS (Breadth-First Search) explores nodes layer-by-layer, searching all neighbors at the current depth before moving deeper. It is complete and guarantees finding the shallowest solution path, but requires high memory ($O(b^d)$ space). DFS (Depth-First Search) explores as deep as possible down a branch before backtracking. It requires much less memory ($O(bm)$ space) but is not guaranteed to find the optimal path and can get stuck in infinite loops.

Question 6

What is Natural Language Processing (NLP)?

Accepted Answer

NLP is a subfield of AI focused on the interaction between computers and human language. The goal is to build systems capable of reading, understanding, processing, and generating human language in a valuable way (tasks include tokenization, parsing, translation, sentiment analysis, and question answering).

Question 7

What is the difference between Tokenization, Lemmatization, and Stemming?

Accepted Answer

1. Tokenization: Splitting a string of text into individual words or characters (tokens).
2. Stemming: A crude heuristic process that chops off the ends of words to reduce them to a base form (e.g. 'running' -> 'run', 'trouble' -> 'troubl').
3. Lemmatization: Uses vocabulary and morphological analysis to return the base dictionary form of a word (e.g. 'better' -> 'good', 'was' -> 'be').

Question 8

What are expert systems in AI?

Accepted Answer

Expert Systems are early AI computer systems (popular in the 1970s and 1980s) that emulate the decision-making ability of a human expert. They consist of a **Knowledge Base** containing structured facts and rules, and an **Inference Engine** that applies logical rules to deduce new facts or solve complex queries.

Question 9

Explain the difference between Deterministic and Stochastic environments in AI.

Accepted Answer

In a Deterministic environment, the next state of the environment is completely determined by the current state and the action performed by the agent (e.g. chess). In a Stochastic (non-deterministic) environment, there is uncertainty—an action can lead to multiple outcomes with differing probabilities (e.g. driving a car).

Question 10

What is an Expert System Inference Engine: Forward Chaining vs Backward Chaining?

Accepted Answer

**Forward Chaining** (data-driven) starts with known facts and applies rules to assert new facts until a goal state is reached. **Backward Chaining** (goal-driven) starts with a goal hypothesis and works backward to find supporting facts or sub-goals from the rules. Forward is active, whereas Backward is investigative.

Question 11

What is Alpha-Beta Pruning?

Accepted Answer

Alpha-Beta Pruning is an optimization technique for the minimax algorithm used in two-player games (like chess). It reduces the search tree by stopping the evaluation of a branch when it finds a move that is guaranteed to be worse than a previously evaluated move, saving substantial compute time without affecting the final minimax decision.

Question 12

What is the difference between Agent and Environment?

Accepted Answer

An **Agent** is an autonomous entity that perceives its environment through sensors and acts upon it through actuators. The **Environment** is everything surrounding the agent, which responds to the agent's actions and provides sensory feedback inputs.

Question 13

What is the Chinese Room Argument?

Accepted Answer

Proposed by John Searle in 1980, the Chinese Room is a thought experiment challenging Strong AI. A person inside a locked room uses a rulebook in English to process Chinese characters received from outside, outputting correct Chinese responses. To outsiders, the person appears to understand Chinese, but they only follow syntax rules without semantic understanding. Searle argues this is how computers operate, demonstrating that syntax is not sufficient for real minds or understanding.

Question 14

What are Genetic Algorithms?

Accepted Answer

Genetic Algorithms are search and optimization heuristics inspired by Charles Darwin's theory of natural evolution. They start with a population of candidate solutions, which undergo selection (fitness scoring), crossover (combining parent solutions), and mutation (introducing random alterations) to evolve optimal solutions over generations.

Question 15

What is the difference between parsing and tagging in NLP?

Accepted Answer

Tagging (e.g. POS tagging) assigns grammatical parts of speech (noun, verb, adjective) to individual words in a sentence. Parsing (syntactic parsing) analyzes the structural relationships between words to build a parse tree, showing how words group into phrases and sentences.

Question 16

Explain the concept of Knowledge Representation.

Accepted Answer

Knowledge Representation is the formal study of how an agent represents real-world facts and information in a computer-readable format (e.g. semantic networks, frames, logical predicates, ontologies) so that inference engines can reason and solve complex problems.

Question 17

What is standard search vs local search in AI?

Accepted Answer

Standard search algorithms (BFS, A*) maintain path histories to trace a sequence from start to goal. Local search algorithms (Hill Climbing, Simulated Annealing) operate on a single current state, modifying it locally to find an optimal solution state, ignoring path history entirely (useful for resource allocation/scheduling).

Question 18

Explain the Hill Climbing algorithm and its limitations.

Accepted Answer

Hill Climbing is a mathematical optimization algorithm that iteratively moves in the direction of increasing value (up the hill) until it finds a peak. Limitations include:
1. Local Maxima: Getting stuck on sub-optimal peaks that are higher than surrounding states but lower than the global optimum.
2. Plateaus: Flat areas where all neighbors have the same value, leaving the algorithm with no directional guidance.

Question 19

What are AI Ethics and Safety concerns?

Accepted Answer

Key concerns include:
1. Bias and Fairness: Models replicating historical prejudices present in training datasets.
2. Transparency: Deep models acting as 'black boxes' whose decisions cannot be explained.
3. Safety and Control: Ensuring autonomous systems (like cars or drones) operate safely.
4. Job Displacement: Automatization impacting workforces.

Question 20

How does TF-IDF work in document information retrieval?

Accepted Answer

TF-IDF (Term Frequency-Inverse Document Frequency) measures how important a word is to a document in a collection. It is the product of:
1. Term Frequency (TF): How often a word appears in a specific document.
2. Inverse Document Frequency (IDF): Penalizes words that appear frequently across all documents (like 'the', 'is'), highlighting unique keywords.

AI Theory & General Interview

Difficulty Filters

Tools & Options

What is Artificial Intelligence, and what is the difference between Weak AI and Strong AI?

Explain the Turing Test.

What is a Heuristic function in AI search algorithms?

Explain the A* Search algorithm.

Explain BFS vs DFS in state space search.

What is Natural Language Processing (NLP)?

What is the difference between Tokenization, Lemmatization, and Stemming?

What are expert systems in AI?

Explain the difference between Deterministic and Stochastic environments in AI.

What is an Expert System Inference Engine: Forward Chaining vs Backward Chaining?

What is Alpha-Beta Pruning?

What is the difference between Agent and Environment?

What is the Chinese Room Argument?

What are Genetic Algorithms?

What is the difference between parsing and tagging in NLP?

Explain the concept of Knowledge Representation.

What is standard search vs local search in AI?

Explain the Hill Climbing algorithm and its limitations.

What are AI Ethics and Safety concerns?

How does TF-IDF work in document information retrieval?

Explore Related Tools

Bulma Box Component – Simple Container with Shadow

Bulma Modal Component – Dialogs, Popups & Overlays

Permutation and Combination Calculator (nPr & nCr)

SQL SELECT Statement – Retrieve Data from Database

Bootstrap 5 Block Buttons: Full

Centimeters to Pixels Converter

Follow Us

Our Tools

Our Company

Special Tools