See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/375488989

The Role of Artificial Intelligence in Video Game Development

Article in Kultura polisa · November 2023

DOI: 10.51738/Kpolisa2023.20.3r.50f

CITATION READS

1 469

1 author:

Aleksandar Filipović
Pravni fakultet za privredu i pravosuđe
58 PUBLICATIONS 159 CITATIONS

SEE PROFILE

All content following this page was uploaded by Aleksandar Filipović on 08 November 2023.

The user has requested enhancement of the downloaded file.

 The Role of Artificial Intelligence in Video
Game Development

Aleksandar Filipović
Faculty of Law for Commerce and Judiciary in Novi Sad

Article Information*·
Research Article • UDC: 004.8:004.928
Volume: 20, Issue: 3, pages: 50–67
Received: September 15, 2023 • Accepted: October 14, 2023
https://doi.org/10.51738/Kpolisa2023.20.3r.50f

Author Note

Aleksandar Filipović https://orcid.org/0000-0002-1097-2079

I have no known conflict of interest to disclose.
Corresponding author: Aleksandar Filipović
E-mail: [email protected]

* Cite (APA):
Filipović, A. (2023). The Role of Artificial Intelligence in Video Game Development.
Kultura polisa, 20(3), 50–67, https://doi.org/10.51738/Kpolisa2023.20.3r.50f

© 2023 by the authors. This article is an open-access article distributed

under the terms and conditions of the Creative Commons Attribution (CC
BY) license (https://creativecommons.org/licenses/by/4.0/).

Page 50 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Abstract
In this paper, we study the vital role of artificial intelligence (AI) in the
development of video games, with a focus on various aspects of AI ap-
plication in this industry. In the introduction, we discuss both the de-
velopment of video games and the role of AI systems in the user
experience, defining the progression of AI’s role in video games. In the
following section, we investigate how in-game entities and AI collabo-
rate. Here, we analyze basic concepts such as Non-Playable Characters
(NPCs) and how AI enhances their intelligence and reactivity in the
game. The mechanisms of AI in video games are a crucial point of con-
sideration in the next part of the paper. We explain how various AI
techniques are used for decision-making, player tracking, and adapting
the game to their actions. Furthermore, we explore the use of AI in
video games beyond NPC control, examining examples such as proce-
durally generated worlds and player experience modeling. This appli-
cation of AI contributes to a deeper and more dynamic player
experience. In the modern application of AI in video games, we delve
into advanced uses of machine learning and deep neural networks in
game development. Here, we consider how AI is used for game per-
sonalization, user data analysis, and enhancing graphics and sound.
Finally, we discuss the future of video games and the role of neural
networks in their development. We predict the growth of AI in various
aspects of games and how it will shape the future player experience.
In conclusion, we assert that artificial intelligence has become an in-
dispensable part of video game development, and its impact will in-
evitably expand in the future, enabling increasingly rich, dynamic, and
personalized games for players worldwide.
Key words: video game entities, video game mechanics, AI
mechanisms in video games, NPCs, neural networks

Page 51 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

The Role of Artificial Intelligence in Video Game

Development

Introduction

Raison d’être of every video game resides deep within the

competitive sphere of human mental constitution, particularly in its
innate impulses (Russell, 2009). A video game is always a competition
with someone or something: either an opponent, a rival, or an adversary.
This competitiveness is sometimes perceived as an “inherent animal
spirit” of humans, which ultimately leads to an “immaterial civil war”
(see more: Lukić, 2010). Given that the real protagonist of a video game
is necessarily homo ludens, that is, a human, either directly or through
their avatar (although the first video games did not have avatars per se,
and interaction between the player and the software occurred directly,
as in the game Tetris), their opponent is always the software that powers
the device on which the game is played, and whose software must adapt
and oppose its activities in the game to compete with and often challenge
or enhance the player’s experience (especially in certain genres where
the goal is not for the player to win but to stay in the game as long as
possible, as seen in earlier games like Space Invaders) (see more:
Filipović, 2022). Thus, even in rudimentary video games, we have the
use of artificial intelligence in shaping gameplay. Individuals who have
experienced video games have unquestionably engaged with artificial
intelligence (AI). Irrespective of the video game genre, AI has
consistently governed various aspects. Typically, this governance is
linked to the conduct of non-player characters, be they neutral or
antagonistic figures.
Software or primary artificial intelligence in video games is
defined as a set of software techniques used in a game engine to
simulate intelligence in the actions of computer-controlled characters
reflecting one of the essential characteristics of intelligence, which is
“perceiving relevant relationships in a specific situation” (Bajac &
Bjelajac, 2022). The game engine that controls AI in video games, in
addition to traditional AI methods, also includes algorithms from control
theory, robotics, computer graphics, and computer science in general.

Page 52 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

The implementation of AI significantly impacts the game, system

requirements, and the game’s budget. Game developers aim to create
interesting and undemanding AI at the lowest cost possible. Artificial
intelligence in video games widely employs various simplifications,
emulations, and even slight deceptions, so the approach and ethics of
using AI in game development differ significantly from the approach to
cognitive artificial intelligence (see more: Bjelajac, Filipović & Stošić,
2022). In a broader sense, artificial intelligence, which we are witnessing
a massive proliferation of applications for, brings forth challenges, risks,
and threats on the internet, especially for children and young individuals.
Consequently, special attention must be paid to the protection of these
vulnerable population groups (see more: Bjelajac & Filipović, 2020;
Bjelajac & Filipović, 2021a). This preventive approach is most effective
when the principles of internet safety culture are applied (see more:
Bjelajac & Jovanović, 2013), alongside the establishment of appropriate
local and international regulations (see more: Bjelajac, Matijašević &
Dimitrijević, 2012). It poses an additional threat for even deeper human
alienation (see more: Bjelajac, 2014). On the other hand, what we refer
to as artificial intelligence in the context of video game creation is
significantly narrower in scope and inherently contained, thus not posing
the same type of risk as narrow artificial intelligence (ANI).
Purists even argue that using the term “artificial intelligence” in
the context of “artificial intelligence in video games” is scientifically
unfounded (Uofa, 2023) since AI in video games does not conform to
the scientific definition of AI and only uses a few branches of academic
“artificial intelligence” science. While “true” AI refers to branches of
self-learning systems and decision-making based on arbitrary data
input, even to the ultimate goal of “cognitive” artificial intelligence
capable of reasoning, AI in video games often consists of a few basic
heuristic rules that are sufficient to provide the player with a good
game, a satisfying experience, and an enjoyable game (Uofa, 2023).

Video game entities and AI

Since the earliest games, players have interacted with

autonomous entities in video games, commonly referred to as non-
playable characters (NPCs). The term NPC was first used for characters

Page 53 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

in role-playing games but is now widely used for characters in almost

all genres. NPCs populate the game world to provide an immersive
gaming experience, and as such, they should be as realistic as possible
in their appearance, movements, dialogues, and decisions. One of the
most overlooked challenges in artificial intelligence (AI) for computer
games is creating in-game characters (NPCs) with human-like behavior
(Uludagli et al., 2023). Contemporary NPCs determine their actions in
different situations using specific decision-making methods, allowing
them to change the current state of the game world.
Entities in video games controlled and managed by software
artificial intelligence (AI) can be diverse and include:
NPC (Non-Playable Characters): These are characters within
the game that are not under direct player control. AI is used to manage
their behavior to make them authentic and responsive to player
interactions.
Opponents and Enemies: AI can control opponents or enemy
entities in the game, determining their tactics, abilities, and reactions
to the player.
Allies: AI can control player allies, providing support and
assistance during the game. This includes combat strategies,
navigation, and other forms of support.
Systemic Elements: AI can manage various systemic
elements in the game, such as dynamic weather conditions,
economies, and more, to create a realistic environment.
Procedurally Generated Entities: In some games, AI can
generate randomly or procedurally created entities as part of the
game, such as enemies, levels, or worlds.
Game Mechanics: Artificial intelligence can also be used to
control various gameplay mechanics, such as physics, animations, and
interactions with objects in the game.
Player Experience Customization: AI can analyze player
behavior and adjust the game to create a personalized experience,
including adjusting difficulty levels or game dynamics.
Player Behavior Analysis: AI can track and analyze player
behavior to develop statistics, recommendations, or in-game
challenges.

Page 54 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Player Assistants: AI entities can serve as guides or assistants

to players, providing them with information, advice, or instructions
during the game.
Interactive Dialogue: AI can control the dialogue between
players and NPCs, allowing for diverse and dynamic conversations in
the game.
Depending on the type of game and its objectives, AI can have
different roles and functions in managing these various entities within
the video game.

The mechanisms of operation of AI in video games

People tend to perceive every interaction with AI as a threat

because they view AI as something superior, surpassing human
capabilities in every field. Such a stance is generally correct (see more:
Bjelajac & Filipović, 2021b). However, in video games, it’s different.
Instead of learning how to defeat human players, artificial intelligence
in video games is designed to enhance the gaming experience of
human players (Lou, 2017). Timur Bukharayev (2009) captures the
role of software AI in video games nicely, stating that the “main task
of AI in a video game is not to defeat the player but to provide a good
experience.”
Dr. Harbing Lou (2017) writes that designers often use tricks to
make NPCs appear intelligent. One of the most commonly used tricks,
known as the Finite-State Machine (FSM) algorithm, was introduced
into video game design in the early 1990s (Lou, 2017). In FSM, the
designer generalizes all possible situations the artificial intelligence
could encounter and then programs a specific reaction for each
situation. Essentially, FSM AI would immediately respond to the
player’s action with its pre-programmed behavior. In shooter games,
artificial intelligence may attack the player but withdraw when the
avatar’s health is dangerously low, leading to a potential “game over.”
In an FSM-based game, a specific character has the capability to
execute four fundamental actions in reaction to potential scenarios:
seeking help, evading, wandering, and taking action. Many well-known
games, such as Battlefield, Call of Duty, and Tomb Raider, include

Page 55 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

successful examples of FSM AI design. Even the turtles in Super Mario

have rudimentary FSM designs (Lou, 2017).
A more advanced method that developers use to enhance the
individualized gaming experience is the Monte Carlo Algorithm (MCTS
or Monte Carlo Tree Search). The MCTS algorithm was created to avoid
the repetition aspect present in the FSM algorithm. The MCTS
algorithm first evaluates all possible moves available to the NPC at a
given moment. Then, for each of these possible moves, it analyzes all
the actions the player could respond with. Afterward, it returns to
assess the NPC’s estimation based on the player’s actions (Cowling et
al., 2012). This artificial intelligence algorithm was used by IBM to
create Deep Blue, the first chess supercomputer, which made history
on May 11, 1997, by defeating world chess champion Garry Kasparov
in a six-game match.
A similar algorithm is used in many strategy games. However,
since there are far more possible moves than in chess, it’s impossible
to consider them all. In such games, the MCTS algorithm will randomly
select some of the possible moves. As a result, NPC actions become
much less predictable for players. This is crucial in games like
Civilization, which have a vast number of options available to the
computer opponent. Building a tree for every possible choice and
scenario would take a very long time. Therefore, to avoid such
extensive calculations, the MCTS algorithm randomly selects a few
possible options. Consequently, the game consumes fewer system
resources while still surprising players with its AI (Lou, 2017).

The use of AI in video games other than NPC control

The use of AI in game development is not necessarily limited to

NPC control. Georgios N. Yannakakis (2012) highlighted, in addition to
NPC control, four potential areas of AI application in video games:
Player Experience Modeling: Recognizing the abilities and
emotional state of the player to adapt the game appropriately. This can
include dynamically balancing the game’s difficulty, adjusting the
game’s difficulty in real-time based on the player’s abilities. AI in
games can also help infer player intentions (such as motion
recognition).

Page 56 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Procedural Content Generation: Creating game environment

elements such as environmental conditions, levels, and even music
automatically. Artificial intelligence methods can generate new content
or interactive stories.
User Behavior Data Mining: This allows game designers to
explore how people use the game, which parts they play the most, and
what causes them to stop playing, enabling developers to fine-tune
the game or improve monetization.
Alternative Approaches to NPC Characters: These involve
changing game settings to enhance the believability of NPCs and
exploring social, rather than individual NPC behavior.
Instead of procedural generation, some researchers (Goodfellow
et al., 2014) have used Generative Adversarial Networks (GANs) to
create new content. Researchers at the University of Cornwall
implemented GANs in 2018 to generate a thousand levels that
developers had written for the video game DOOM (1993). After
training, the neural network prototype could design new levels for the
game autonomously. Similarly, researchers at the University of
California created a GAN prototype for generating levels for Super
Mario, and in 2020, Nvidia demonstrated a GAN clone of the cult game
Pac-Man. GANs learned how to recreate the game by observing 50,000
game levels, mostly generated by bots.

Contemporary application of AI in video games

Heuristic artificial intelligence algorithms in games are used

across a wide spectrum in many industries within the video game
industry (Korotkov, 2017). The most obvious use of AI in games is in
controlling NPCs, although scripting is also a common form of control.
Pathfinding is another widespread use of artificial intelligence in video
games, especially in real-time strategy games. Pathfinding is a method
for determining how an NPC should move from one point on the map
to another, considering the terrain, obstacles, and possible fog of war.
AI in video games is also associated with dynamic game balancing.
The concept of AI unpredictability has been explored in games such as
Creatures, Black & White, Nintendogs, and toys like Tamagotchi
(Korotkov, 2017). The “pets” in these games have the ability to “learn”

Page 57 of 281
 Kultura polisa
20(3), 50–67

from the player’s actions, and their behavior changes accordingly.

While these solutions are drawn from a limited set of possible solutions,
they often provide the desired illusion of intelligence on the other side
of the screen.
Artificial intelligence (AI) has been a part of computer games for
decades, but recent advances in machine learning and deep learning
have ushered in a new era of “Generative Artificial Intelligence (GenAI).”
This technology is poised to revolutionize the gaming industry by offering
highly impressive, personalized, and realistic gaming experiences
(Gartner, n.d.). One innovative application of artificial intelligence in
games is the creation of hyper-realistic worlds and non-player characters
(NPCs). Game creators can use generative artificial intelligence to train
voice actors and generate dialogue in a fraction of the time it would take
to write scripts by hand. This not only saves time but also allows for the
creation of endless worlds populated by realistic characters. These worlds
can be customized and transformed based on player actions, opening up
endless possibilities for replayability. Another area where AI shines is
dynamic storytelling. Traditionally, game narratives have been devised
by developers, but generative AI enables unplanned narrative
digressions and personalized stories. By analyzing player behavior and
choices, artificial intelligence in video games can adapt and offer highly
personalized stories while maintaining the overall game structure
(Gartner, n.d).
GenAI also opens up new possibilities for creating levels,
missions, and side tasks. Game developers can use generative artificial
intelligence to create unique missions or levels based on player skills,
progress, or characteristics. This not only adds diversity to the game
but also retains player interest. Music generation is another area where
AI has an impact. AI-based tools can compose original music for video
games based on user input. With further advances in AI, in the future,
it will be able to create music that matches the style and mood created
by game developers.
One of the most significant applications of AI in games is real-
time image enhancement (Crider, 2023). Technologies like Nvidia Deep
Learning Super Sampling (DLSS), AMD Fidelity Super Resolution (FSR),
and Intel Xe Super Sampling (XeSS) have already improved graphical
fidelity in video games by increasing frame rates while maintaining

Page 58 of 281
 Kultura polisa
20(3), 50–67

graphics quality. As AI continues to advance, real-time image

enhancement will continue to evolve, offering players stunning visual
experiences (Archer, 2022).
AI has an essential role in the development and testing of video
games. It can automate tedious tasks such as bug detection and
gameplay testing, saving time and developer resources. AI enhances
game testing, a critical aspect of game development. Traditional game
testing, involving human testers or scripted bots, has its limitations.
AI-based testing addresses these limitations by automating many
aspects of game testing, reducing the need for human testers and
speeding up the process (Hawk, 2023).
Predictive analysis using artificial intelligence algorithms can
help identify potential issues before they become serious, providing
players with a smoother gaming experience. AI can be used to improve
storytelling in video games by helping to generate narrative content
or enhance the overall storytelling experience. Natural language
processing (NLP) techniques can be used to analyze player feedback
and adapt the narrative in response, generate new content including
unique character backgrounds, create new dialogue options, or even
generate new stories (Hawk, 2023). AI is ready to revolutionize the
gaming industry with hyper-realistic worlds, dynamic storytelling,
personalized experiences, and improved game development processes.
As this technology continues to evolve, players can expect even more
impressive and enjoyable gaming experiences in the future.
Gartner (n.d) sees “Generative Artificial Intelligence becoming
a general-purpose technology with an impact akin to that of the steam
engine, electricity, and the internet. The hype will settle as the reality
of implementation takes hold, but the impact of generative artificial
intelligence will continue to grow as people and enterprises discover
more innovative applications for the technology in everyday work and
life.”

Neural networks and the future of gaming

Neural networks, specifically artificial neural networks (ANN) or

simulated neural networks (SNN), constitute a subset of machine
learning and serve as the backbone of deep learning algorithms. They

Page 59 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

are called “neural” because they mimic the way neurons in the human
brain communicate with each other (IBM Data and AI Team, 2023).
Possessing the capacity to acquire knowledge from data and formulate
forecasts, they are exceptionally well-suited for tasks like image and
speech recognition, processing natural language, and autonomously
making decisions.
The utilization of neural networks in video games holds the
promise of transforming the way games are crafted, produced, and
experienced. Neural networks can be employed to create smarter and
more advanced artificial intelligence systems for non-player characters
(NPCs) in games. NPCs powered by neural networks will be able to
react to players’ actions in a more sophisticated and realistic way,
making gameplay more impressive and enjoyable (Kagan et al., 2022).
Neural networks can also enhance the player’s experience in
other ways. They can be used to create procedurally generated
content, such as levels, enemies, and unique weapons tailored to each
player. This means that each player will have a unique gaming
experience, and no two playthroughs will be exactly the same.
Another application of neural networks in games is
reinforcement learning (Jones, 2017), a subfield of machine learning
that focuses on training systems to make decisions based on rewards
and penalties. Reinforcement learning is applied to instruct neural
networks to excel in games like chess and Go, surpassing human
capabilities, and it holds the potential to achieve similar results in
diverse game genres, including first-person shooter games and real-
time strategy games.
In most modern video games, NPC behavior is scripted; no
matter how many times a player exploits an NPC’s weakness, that
weakness is never improved. However, if NPCs in a game could learn
through interaction with the player, their behavior could improve over
the course of gameplay, keeping the game interesting. The
NeuroEvolution of Augmenting Topologies (rtNEAT) method is used to
develop increasingly complex artificial neural networks in real-time
while the game is being played (Stanley et al., 2005). The rtNEAT
technique permits agents to adapt and enhance their performance as
the game progresses. In essence, rtNEAT has the potential to usher in
a novel category of video games where players instruct a group of

Page 60 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

agents through a sequence of tailored training drills. To illustrate this

concept, in the NeuroEvolvingRoboticOperatives (NERO) game, players
train a squad of robots for combat. Looking ahead, rtNEAT could open
up possibilities for fresh educational and training applications that
adjust in real-time as users acquire new skills.
Finally, neural networks can be used to create adaptive games
(Redepenning et al., 2022) that change and evolve based on player
behavior to enable individuals with disabilities to actively participate in
gaming communities. Current adaptive gaming technology not only
allows individuals to engage in games but also to compete on a level
playing field with those who do not use adaptive gaming equipment.
Games based on neural networks can generate unique challenges for
each player based on their motor skills and gameplay style and can
adjust the game’s difficulty level in real-time based on the player’s
progress. This level of adaptability can lead to a more personalized and
enjoyable gaming experience for an underserved population of players.
Studies (Redepenning et al., 2022) show clear evidence of improved
quality of life and social benefits from participating in adaptive gaming,
strongly supporting the need to increase the availability of adaptive
gaming services as part of the rehabilitation of individuals with
disabilities.
Current applications of neural networks for improving the
gaming experience suggest that rtNEAT technology brings immediate
commercial value to modern games. Any game in which NPC behavior
endlessly repeats and becomes boring can be improved by allowing
rtNEAT to partially modify tactics in real-time. This is particularly useful
in persistent video games like Massive Multiplayer Online Games
(MMOGs) that last for months or years. In these games, rtNEAT’s
potential is constantly adapted and optimized for NPC behavior,
potentially permanently altering the gaming experience for millions of
players worldwide (Stanley et al., 2005).
Incorporating neural networks into video games can inaugurate
a fresh era of gaming, offering players more immersive and tailored
experiences.. As technology continues to evolve and improve, it is
exciting to think about what the future of gaming may look like and
what role neural networks will play in shaping it.

Page 61 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Conclusions

It’s incredible how much artificial intelligence has achieved in

so many industries in such a relatively short time. However, the
question arises: what do we do next? Will artificial intelligence have
the capability to completely replace humans within a few decades? Or
are there things that humans are capable of doing that AI will never
learn? In any case, many good things can come out of this. Although
we can’t know exactly what will happen, we are excited to see what
the future of artificial intelligence will bring.
Artificial intelligence is an swiftly progressing technology, and its
possible uses in game design are striking. Game design with artificial
intelligence could potentially transform the gaming industry by
facilitating the development of hyper-realistic virtual environments and
customizable difficulty levels. However, there are still many challenges
to overcome before these potentials can be realized. Nevertheless, as
artificial intelligence technology continues to improve, it is likely that AI
game design will become an integral part of future games.
Video games have come a long way from Pong and Pac-Man,
and with the advent of artificial intelligence, game developers can now
create virtual worlds where anything is possible. But with this new
technology, what kind of future can we realistically expect in terms of
game design?
The potential of AI game design is impressive, but there are still
several significant challenges to overcome before it can be realized.
One of the biggest obstacles is cost. Designing games with artificial
intelligence requires immense computational power, which can be quite
expensive. Additionally, ethical and legal considerations need to be
addressed, as well as potential questions about the “soul” or “mind” of
AI-driven games. The development of AI game design should also be
accompanied by advancements in artificial intelligence algorithms. AI
algorithms are still relatively simple, and they need to be improved to
unlock the technology’s full potential. As AI becomes more
sophisticated, more complex interactions and responses can be
implemented, making games even more enjoyable. It is clear that the
use of artificial intelligence in game design has the potential to
revolutionize the gaming industry. AI game design has the potential to

Page 62 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

create an entirely new type of game where players are truly immersed
in virtual worlds. This technology could also enable the creation of
games that continually self-improve, increasing their capacity.
However, there are still many challenges to overcome before all
these ideas become applicable. More research is needed on the
development and implementation of AI game design, taking into
account ethical and legal implications. It is evident that AI game design
is still in its infancy, and only time will tell what kind of games will
emerge in the future. In the end, “to be useful, AI must fight well but
lose more often than it wins. Such a role should make the player feel
smart, cunning, and powerful” (Buckland, 2004).
In the future, the development of artificial intelligence in games
will likely not focus on creating more powerful NPCs that seek
sophisticated ways to defeat players (Lebedev, 2023). Instead, the
focus will be on how to create a unique player experience for each
individual. Players today pay great attention to details, including not
only the appearance and quality of graphics but also how alive and
interactive the game is in every possible way. Artificial intelligence is
what can elevate the gaming experience to a higher level. Perhaps one
day, players won’t be able to tell whether a character in the game is
controlled by artificial intelligence or another player.

References

Archer, J. (2022, January 17). Nvidia DLDSR tested: better visuals and
better performance than DSR. Rock Paper Shotgun. https://
www.rockpapershotgun.com/nvidia-dldsr-tested-better-visuals-
and-better-performance-than-dsr
Bajac, M., & Bjelajac, Ž. (2022). Artificial Intelligence (AI) in the Function
of Recognizing Emotions and Violent Mentality. Kultura polisa,
19(2), 277–297. https://doi.org/10.51738/Kpolisa2022.19.2p.
277bb
Bjelajac, Ž., Matijašević, J., & Dimitrijević, D. (2012). Značaj
uspostavljanja međunarodnih standarda u suzbijanju
visokotehnološkog kriminala [The Significance of Establishing

Page 63 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

International Standards in Combating High-Tech Crime].

Međunarodna politika, 63(1146): 66–84.
Bjelajac, Ž., & Jovanović, M. (2013). Pojedini aspekti bezbednosne
kulture na Internetu [Some Aspects of Security Culture on the
Internet]. Kultura polisa, 10(21), 99–114.
Bjelajac, Ž. (2014). Alijenacija čoveka u eri naučnotehnološkog razvoja
i globalizacije kao uzrok porasta i raznolikosti organizovanog
kriminala [The alienation of man in the era of scientific and
technological development and globalization as a cause of the
increase and diversity of organized crime]. Kultura polisa, 11(25):
385–396.
Bjelajac, Ž., Filipović, A. (2020). Perspektive zaštite dece od zloupotreba
na internetu [Perspectives of child protection from Internet
misuse]. Kultura polisa, 17(41), 259–271.
Bjelajac, Ž., & Filipović, A. (2021a) Fleksibilnost digitalnih medija za
manipulativno delovanje seksualnih predatora [Flexibility of
digital media for manipulative activities of sexual predators].
Kultura polisa, 18(44), 51–67.
Bjelajac, Ž., Filipović, A. (2021b). Artificial Intelligence: Human Ethics in
Non-Human Entities. In Proceedings of the 3rd Virtual
International Conference „Path to a Knowledge Society-Managing
Risks and Innovation – PaKSoM 2021”. Mathematical Institute of
Serbian Academy of Sciences and Arts.
Bjelajac, Ž., Filipović, A., & Stošić, L. (2022). Quis custodiet ipsos
custodes: ethical dilemmas of the KM governed by AI. In
Proceedings of the 4th Virtual International Conference „Path to a
Knowledge Society-Managing Risks and Innovation – PaKSoM
2022”. Mathematical Institute of Serbian Academy of Sciences and
Arts.
Buckland, M. (2004). Programming Game AI by Example, Jones &
Bartlett Learning.
Bukharaev, T. (2009, May 28). Iskusstvennyy intellekt v Heroes of Might
and Magic V [Artificial Intelligence in Heroes of Might and Magic
V], Gamer. http://www.gamer.ru/heroes-of-might-and-magic-v-

Page 64 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

poveliteli-ordy/iskusstvennyy-intellekt-v-heroes-of-might-and-
magic-v
Cowling, P. I., Powley, E. J., & Whitehouse, D. (2012). Information set
Monte Carlo Tree search. IEEE Transactions on Computational
Intelligence and AI in Games, 4(2), 120–143.
https://doi.org/10.1109/tciaig.2012.2200894
Crider, M. (2023, August 24). Upscaling Explained: DLSS vs FSR vs
XeSS. PCWorld. https://www.pcworld.com/article/2041595/
upscaling-explained-nvidia-dlss-vs-amd-fsr-vs-intel-xess.html
Filipović, A. (2022). Bića i svetovi video igre [Beings and Worlds of Video
Game]. Pravni fakultet za privredu i pravosuđe.
Gartner. (n.d.). Generative AI: What is it, tools, models, applications and
use cases. https://www.gartner.com/en/topics/generative-ai
Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D.,
Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial
networks. arXiv (Cornell University). https://doi.org/10.48550
/arxiv.1406.2661
Hawk. (2023, June 7). AI in Video Game Development: A
Comprehensive Guide. Hawk. https://hawk.live/posts/ai-impact-
video-game-development
IBM Data and AI Team. (2023). AI vs. Machine Learning vs. Deep
Learning vs. Neural Networks: What’s the difference? IBM Blog.
https://www.ibm.com/blog/ai-vs-machine-learning-vs-deep-
learning-vs-neural-netw
Jones, M. T. (2017, October 10). Train a software agent to behave
rationally with reinforcement learning. IBM Developer.
https://developer.ibm.com/articles/cc-reinforcement-learning-
train-software-agent/
Kagan, B. J., Kitchen, A. C., Tran, N. T., Habibollahi, F., Khajehnejad, M.,
Parker, B. J., Bhat, A., Rollo, B., Razi, A., & Friston, K. J. (2022).
In vitro neurons learn and exhibit sentience when embodied in a
simulated game-world. Neuron, 110(23), 3952-3969.e8. https://
doi.org/10.1016/j.neuron.2022.09.001

Page 65 of 281
THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Korotkov, T. (2017, November 3). Igrovoy iskusstvennyy intellekt

[Gaming Artificial Intelligence]. Prezi. https://prezi.com/p/
p48me7jkpwbh/presentation/?fallback=1
Lebedev, V. (2023). Ne sovsem chelovek: Iskustvennyy intellekt v igrah
[Not quite human: artificial intelligence in games]. skillbox.ru.
https://skillbox.ru/media/gamedev/iskusstvennyy-intellekt-v-
igrakh/
Lou, H. (2017, August 28). AI in Video Games: Toward a More Intelligent
Game – Science in the News. Science in the News.
https://sitn.hms.harvard.edu/flash/2017/ai-video-games-
toward-intelligent-game/
Lukić, K. (2010). Komodifikacije igre – kritičke perspektive umetnosti
digitalnih igara, Kultura, 127, 236-251.
Redepenning, D. H., Huss, S. A., & Maddali, S. (2022). Influence of
adaptive video gaming on quality of life and social relationships.
Assistive Technology, 35(4), 339–346. https://doi.org/10.1080/
10400435.2022.2067913
Russell, B. (2009). Authority and the individual. Routledge.
Stanley, K. O., Bryant, B. D., & Miikkulainen, R. (2005). Evolving Neural
Network Agents in the NERO Video Game (2005). In Proceedings
of the IEEE 2005 Symposium on Computational Intelligence and
Games (CIG’05). IEEE. https://nn.cs.utexas.edu/?stanley:cig05
Uludağlı, Ç., & Oguz, K. (2023). Non-player character decision-making
in computer games. Artificial Intelligence Review. https://doi.org/
10.1007/s10462-023-10491-7
Uofa. (2022, January 7). Directions and areas of application of artificial
intelligence. Map of the application of artificial intelligence
technologies, uofa.ru, https://uofa.ru/en/napravleniya-i-oblasti-
primeneniya-iskusstvennogo-intellekta-karta/
Yannakakis, G. N. (2012). Game AI revisited. In Proceedings of the 9th
conference on Computing Frontiers (CF ’12). Association for
Computing Machinery. https://doi.org/10.1145/2212908.2212954

Page 66 of 281
 THE ROLE OF AI IN VIDEO GAME DEVELOPMENT Kultura polisa
Aleksandar Filipović 20(3), 50–67

Uloga veštačke inteligencije u kreiranju video igara

Aleksandar Filipović
Pravni fakultet za privredu i pravosuđe u Novom Sadu

Sažetak
U ovom radu istražujemo ključnu ulogu veštačke inteligencije (AI) u
razvoju video igara, sa fokusom na različite aspekte primene AI u ovoj
industriji. U uvodu, govorimo kako o razvoju video igara, tako i o ulozi AI
sistema u korisničkom iskustvu, definišući progresiju uloge AI u video
igrama. U sledećem delu, istražujemo kako entiteti video igara i AI
sarađuju. Ovde analiziramo osnovne koncepte kao što su NPC (Non-
Playable Characters) i kako AI unapređuje njihovu inteligenciju i
reaktivnost u igri. Mehanizmi delovanja AI u video igrama su ključna tačka
razmatranja u narednom delu rada. Objašnjavamo kako se različite
tehnike AI koriste za donošenje odluka, praćenje igrača i prilagođavanje
igre njihovim akcijama. Dalje, istražujemo upotrebu AI u video igrama
izvan kontrole NPC-a, istražujući primere kao što su proceduralno
generisani svetovi i modeliranje iskustva igrača. Ova primena AI doprinosi
dubljem i dinamičnijem iskustvu igrača. U savremenoj aplikaciji AI u video
igrama, istražujemo napredne primene mašinskog učenja i dubokih
neuronskih mreža u razvoju igara. Ovde razmatramo kako se AI koristi za
personalizaciju igara, analizu korisničkih podataka i unapređenje grafike i
zvuka. Na kraju, raspravljamo o budućnosti video igara i ulozi neuronskih
mreža u njihovom razvoju. Predviđamo rast AI-a u raznim aspektima igara
i kako će to oblikovati buduće iskustvo igrača. Zaključujemo da je veštačka
inteligencija postala nezaobilazan deo razvoja video igara i da će njen
uticaj samo rasti u budućnosti, omogućavajući sve bogatije, dinamičnije i
personalizovanije igre za igrače širom sveta.
Ključne reči: Veštačka inteligencija (AI), video igre, entiteti video
igre, mehanizmi AI u video igrama, NPC, neuronske mreže.

Page 67 of 281

View publication stats