Game AI Design for Game Planners: Essential Knowledge for Efficient Collaboration with Programmers

Were you afraid to bring up AI in front of programmers?

'How should I design enemy patterns? What is a behavior tree?'
As a game designer, you've probably experienced at least once the frustration of needing to design enemy AI but feeling stuck due to a lack of development knowledge. You want to explain the exciting combat you have in mind, but you often find yourself at a loss for words because you can't establish the 'criteria for decision-making.'

This course is prepared just for you.

We remove difficult coding terminology and discuss AI thoroughly in 'the planner's language'. We explain the principles of core enemy AI elements such as detection, patterns, behavior, and difficulty in a very easy-to-understand way, showing how they should look and function in the game.

Especially don't worry about concepts that sound headache-inducing just from their names, like FSM (Finite State Machine), Behavior Trees, and Utility Systems. Even if you don't know programming, I'll help you understand how they work so you can slap your knee and say "Ah, that's how it works!" and make design decisions on your own.

From action, RPG, FPS to open world.
Instead of telling programmers "Please figure it out," become a practical game designer who communicates clearly by saying "Please make it like this".

You will learn the following content

Section 1. Basics of Enemy AI and Player Recognition

First, from a planner's perspective, how enemy AI operates and reacts we establish that framework.
"Did the enemy see me?", "Is it chasing me now?" The tension players feel ultimately begins with the AI's 'detection' and 'decision-making'. From setting detection ranges to state transitions (Idle→Chase), natural movement and animation connections, we solidify the fundamental logic that makes AI behave as if it's alive.

Section 2. Fairness, Intent Communication, and Difficulty System

How do you create 'good difficulty' that allows players to accept "Ah, I died because I wasn't good enough"?
You'll understand the difference between unfair deaths and challenging fun, and learn design methods that consider telegraphing techniques that signal attacks in advance and human reaction time. Through this, you'll be able to design fair and rhythmic combat where players and enemies exchange moves.

Section 3. AI Constraints and Infrastructure That Planners Need to Understand

Planners don't need to directly code FSM (Finite State Machine) or Behavior Trees. However, understanding their structure changes your design approach.
This helps you understand the logic behind enemy AI decision-making and consider game performance and consistency issues. Through this section, you'll be able to write design documents that programmers welcome as "clear to implement".

Section 4. Advanced Decision-Making Models and Extended Concepts

Beyond simple pattern repetition, how do you design high-intelligence AI that makes its own decisions and formulates strategies based on situations?
We'll explore advanced models used in modern games, including Goal-Oriented Action Planning (GOAP), utility-based selection (scoring), and group AI where enemies cooperate with each other. Instead of complex technical jargon, we'll help you develop the design insight to determine "Which AI model would be most effective for our game?"

Section 5. AI Design Tips That Help in Practice

Beyond theory, a collection of useful practical tips for creating high-quality enemies for actual games.
It covers how to create diversity by assembling patterns like blocks, how to create tension by controlling cooldowns and randomness, and how to maximize impact through attack and hit effects. Finally, through testing and polishing the AI to make it more fun, you'll complete practical planning capabilities ready for real-world deployment.

Recommended for

• I'm assigned to enemy AI, but I'm a beginner game designer who doesn't know where to start with the design

• Indie/Solo developers who are struggling with monotonous and boring enemy AI patterns

• FSM, Behavior Trees, Utility, and other AI design methods for planners who want to properly understand them from a planning perspective

• Elements like enemy telegraphing, fairness, and difficulty adjustment are system designers who find it hard to explain why things aren't working well

• Developers struggling to create a consistent combat experience because combat flow and AI behavior are disconnected

After taking the course, you'll see these changes

• You will be able to design enemy AI based on logical, systematic principles and theories rather than rough intuition.

• You will develop criteria to analyze and explain why enemies are boring or unfair in your current project.

• You can comprehensively design the elements that make up enemy AI, such as detection, patterns, behavior, and difficulty levels.

• You will develop the ability to select and apply appropriate AI models such as FSM, Behavior Trees, GOAP, and Utility AI according to game situations.

• From pattern design to tactical AI, randomness control, and enhanced direction, you can directly create and improve high-quality enemy AI.

Features of This Course

All lecture slides are provided as PDF files.

The slides used in the lecture are provided as PDF materials for review.
By utilizing these materials, you can take notes directly on the PDF or print them out for note-taking,
and refer to them frequently when designing your own projects or portfolios.

Clear explanations using AI voice and condensed progression enable efficient learning.

Like my other courses, this course was also recorded using AI voice. Since I used the latest AI model, it delivers the course content with accurate pronunciation and natural explanations that rival professional voice actors.

All lectures were recorded and edited using pre-prepared lecture scripts, resulting in the most condensed lectures with not a single second of wasted time.

Additionally, thanks to AI's accurate pronunciation, nearly 100% accurate automatic subtitles are generated. Therefore, it has become possible to deliver improved lecture content to those who watch classes with subtitles enabled.

And I've added slight variations to the AI voice for each lecture. Therefore, you can listen without getting bored, as if each lecture is being explained by a different instructor.

Notes Before Enrollment