How AI plays games quiz

Question 1

What does it mean when we say a computer program is a 'game-playing AI'?

Accepted Answer

A program that decides its own moves to try to win a game

Answer

A program that only stores the rules of a game

Answer

A program that draws the game's graphics

Answer

A program that connects players over the internet

Question 2

In games like chess, AI often imagines many future moves before deciding. What is this 'looking ahead' technique generally called?

Accepted Answer

Search (exploring a tree of future moves)

Answer

Rendering

Answer

Compression

Answer

Encryption

Question 3

What is a 'game tree' in AI?

Accepted Answer

A diagram of possible moves and the positions they lead to

Answer

A list of the game's high scores

Answer

The shape of the game board

Answer

A type of computer memory

Question 4

In a two-player board game, what is the Minimax algorithm trying to do?

Accepted Answer

Pick the move that is best assuming the opponent plays their best too

Answer

Pick a random legal move

Answer

Always copy the opponent's last move

Answer

Pick the move that ends the game fastest

Question 5

Why can't a chess AI simply check every possible future move all the way to the end of the game?

Accepted Answer

There are far too many possible move sequences to count them all

Answer

Chess rules change too often

Answer

Computers cannot understand chess rules

Answer

The board is too small

Question 6

What is an 'evaluation function' in a game-playing AI?

Accepted Answer

A way to score how good a position is for the player

Answer

A way to count how long the game has lasted

Answer

A list of the rules of the game

Answer

A way to choose the game's colors

Question 7

What is a 'heuristic' in the context of AI playing games?

Accepted Answer

A helpful rule of thumb that guides good guesses

Answer

A guaranteed perfect strategy

Answer

A type of game controller

Answer

A way to cheat at the game

Question 8

Alpha-beta pruning is used with Minimax. What problem does it help solve?

Accepted Answer

It skips branches that cannot change the result, saving time

Answer

It makes the AI play randomly

Answer

It changes the rules of the game

Answer

It adds more branches to search

Question 9

What is reinforcement learning, and why is it useful for games?

Accepted Answer

Learning by trial and error using rewards and penalties

Answer

Memorizing a fixed list of moves from a book

Answer

Copying exactly what a human types

Answer

Drawing the game faster

Question 10

In reinforcement learning, what is a 'reward'?

Accepted Answer

A signal telling the AI how good or bad its action was

Answer

The name of the game's main character

Answer

The graphics quality setting

Answer

A list of the game's rules

Question 11

DeepMind's AlphaGo famously beat a top human player at which board game in 2016?

Accepted Answer

Go

Answer

Monopoly

Answer

Checkers

Answer

Snakes and Ladders

Question 12

AlphaZero learned to play chess and Go without studying human games. How did it learn?

Accepted Answer

By playing millions of games against itself (self-play)

Answer

By memorizing a famous chess book

Answer

By asking grandmasters for advice

Answer

By watching online chess videos

Question 13

What is 'self-play' and why does it help an AI improve?

Accepted Answer

The AI plays against copies of itself to keep improving

Answer

The AI plays alone with no opponent

Answer

A human plays against the AI repeatedly

Answer

The AI watches recorded games

Question 14

How does a neural network help a game-playing AI like AlphaGo?

Accepted Answer

It estimates good moves and position strength to guide the search

Answer

It stores the game's sound effects

Answer

It draws the game board on screen

Answer

It connects players online

Question 15

What does 'training data' mean for an AI that learns from past games?

Accepted Answer

The example games the AI studies to learn patterns

Answer

The score the player earns

Answer

The speed of the computer

Answer

The game's background music

Question 16

In a simple grid game, an AI must balance trying new paths versus using paths it knows are good. What is this trade-off called?

Accepted Answer

Exploration versus exploitation

Answer

Minimax versus Maximin

Answer

Win versus lose

Answer

Online versus offline

Question 17

Why is tic-tac-toe considered a 'solved' game for AI?

Accepted Answer

Every possible move can be fully analyzed, so perfect play is known

Answer

It has no rules

Answer

Only humans can win it

Answer

It uses no strategy at all

Question 18

What is a 'state' in a game from an AI's point of view?

Accepted Answer

A complete snapshot of the game at a single moment

Answer

The player's username

Answer

The brand of the computer

Answer

A bug in the game

Question 19

How do AI opponents (bots) in many video games often decide their behavior?

Accepted Answer

By following rules and scripted behaviors designed by developers

Answer

By reading the player's mind

Answer

By randomly guessing every frame with no logic

Answer

By connecting to a human controller

Question 20

A game-playing AI sometimes uses random rollouts to estimate a move's value. This method is part of which famous search technique?

Accepted Answer

Monte Carlo Tree Search

Answer

Bubble sort

Answer

Linear regression

Answer

Binary search

Question 21

What is the difference between a game with 'perfect information' and 'imperfect information'?

Accepted Answer

Perfect info means everything is visible; imperfect info hides some details

Answer

Perfect info means the AI never loses

Answer

Imperfect info means the game has bugs

Answer

They mean the same thing

Question 22

Why is poker harder for AI than chess in one important way?

Accepted Answer

Players' cards are hidden, so the AI must handle uncertainty and bluffing

Answer

Poker has more pieces than chess

Answer

Poker has no rules

Answer

Poker is played faster

Question 23

What is an 'agent' in AI game terminology?

Accepted Answer

The AI entity that observes the game and takes actions

Answer

The human watching the game

Answer

The screen showing the game

Answer

The game's instruction manual

Question 24

In reinforcement learning, what is a 'policy'?

Accepted Answer

A rule that tells the agent which action to take in each state

Answer

The list of the game's high scores

Answer

The rulebook printed by the game maker

Answer

The amount of memory the game uses

Question 25

How did early AI learn to play Atari video games like Breakout just from the screen pixels?

Accepted Answer

By learning from screen pixels and score using a neural network with reinforcement learning

Answer

By reading the game's source code

Answer

By being told every correct move by a human

Answer

By copying another running game

Question 26

Below is a simple Python idea for picking a move. What is the program doing?

Accepted Answer

Trying each move, scoring the result, and keeping the best-scoring move

Answer

Choosing a completely random move

Answer

Counting how many players are online

Answer

Saving the game to a file

Question 27

Why might a game company NOT want its AI opponents to play perfectly against players?

Accepted Answer

A too-perfect AI would frustrate players and make the game less fun

Answer

Perfect AI is against the law

Answer

Perfect AI uses no electricity

Answer

Players prefer to always lose

Question 28

What is 'difficulty scaling' or 'dynamic difficulty' in game AI?

Accepted Answer

Adjusting the AI's strength based on the player's performance

Answer

Making the screen brighter

Answer

Changing the game's language

Answer

Increasing the download size

Question 29

Sometimes a learning AI finds a weird trick that scores points but breaks the spirit of the game. What is this behavior often called?

Accepted Answer

Reward hacking (exploiting a loophole in the reward)

Answer

Fair play

Answer

Memory leaking

Answer

Pixel rendering

Question 30

Is it cheating when a game's AI uses information a human player cannot see, like the whole map?

Accepted Answer

It can be acceptable for difficulty, but fairer AI uses the same info as humans

Answer

It is always illegal

Answer

It makes the AI weaker

Answer

It only happens by accident

Test yourself