Gaming is a fast-growing industry. Global gaming revenue in 2023 is estimated to be around $221.6 billion, surpassing the combined revenues of movies, music, and books. As the industry expands, it has become a popular playground for artificial intelligence (AI) algorithms, with AI gaming trends shaping the sector's future.
AI and gaming share a strong connection. Games are testing grounds for various AI algorithms developed by companies like DeepMind (StarCraft, Go, Atari games) and OpenAI (DOTA 2). The large datasets generated by games are crucial for training AI algorithms. These AI algorithms are used to create new games, improve existing ones, and help players improve their skills.
The intersection of AI and gaming has attracted the attention of major tech companies, with Netflix, Apple, Microsoft, Amazon, and others investing heavily in the gaming industry. Game executives are also increasingly focusing on AI, with many studios and publishers building AI teams or initiating proof-of-concepts.
In this post, we will explore the different game categories, their business models, and the key areas of AI applications in gaming, divided into content-based and analytics-based focuses. We will also discuss the potential impact of AI on the gaming industry and address concerns surrounding AI's role in the creative process.
Exploring AI in Games: A Comprehensive Industry Overview
The gaming world is vast, with many game categories and genres. While genres refer to the style of gameplay (e.g., battle, shooter, action-adventure, puzzle), game categories are based on monetization models and business focuses. The main game categories are hypercasual, casual, midcore, and core, each with potential AI use cases.
Distinguishing between game categories can be challenging, as the lines between them are unclear. For example, Candy Crush might be considered hypercasual by some and casual by others, while Clash of Clans may be classified as either midcore or casual. Understanding game categories is crucial for identifying relevant AI gaming trends and applications despite these ambiguities.
For instance, designers use AI to create assets, maps, levels, and smarter AI NPCs in core and midcore games to improve narratives and gameplay complexity.
Use Cases by Game Category for Developers
In this section, we’ll explore different game categories and their associated AI use cases, focusing on core, casual, hypercasual, and midcore games.
An Overview of AI in Core Games
Core games like Hogwarts Legacy, Zelda, Final Fantasy, and Grand Theft Auto are complex and require players to invest significant time to master and advance through them. These pay-to-play games rely on amazing content, great game mechanics, complex characters, and aggressive marketing to justify their high upfront costs.
Due to their focus on content, core games are more likely to leverage AI for content generation, creating more levels, assets, better art, and smarter AI characters.
A Look at Use Cases of AI in the Casual Game Category
Casual games, like Candy Crush and Clash of Clans, emerged with the rise of smartphones. These games are easy to play, have simple content, and can be enjoyed in short sessions.
Casual games are free and monetized through in-app purchases (IAP) and ads. Balancing ads and IAP is crucial, as the longer a player engages with the game, the more likely they are to make IAP purchases.
Using AI for video game analytics is essential for casual games. They allow for smarter player segmentation, churn analysis, dynamic pricing, and game personalization to keep users engaged.
Hyper-Casual Game Category Use Cases
Hypercasual games, like Flappy Bird, are even simpler than casual games, with minimalistic UI and content. Players typically engage in hypercasual games for a short period before moving on.
Hypercasual monetization relies almost entirely on ads, with publishers acquiring users through paid ads and monetizing through in-game ads. User acquisition is crucial for hypercasuals, focusing on finding new channels and creating attractive ads to increase install rates.
As the cost per install rises, many hypercasual developers want to incorporate IAP and keep players engaged longer, making AI for analytics use cases more pressing.
Exploring Midcore Game Category Applications
Midcore games like Diablo Immortal, Genshin Impact, and Roblox combine core games' rich content and gameplay with simplified mechanics and shorter session times than casual games. These games are monetized through IAP and ads, making them suitable for AI gaming content generation and analytics use cases.
As game developers continue to innovate and blur the lines between categories, AI gaming trends will play an increasingly important role in shaping the industry's future.
AI Use Cases in the Gaming Industry
AI is transforming gaming content, player engagement, and game development. Let's explore some of the most exciting AI gaming use cases in the video game industry.
AI for Content Generation
The gaming industry is excited about how AI could change content creation. From AAA studios like Bethesda (Skyrim) and Rockstar Games (GTA V) to indie developers, AI-generated assets, create smarterAI NPCs and streamline game development.
Games like Call of Duty and EA's franchises are exploring AI-generated concept art, level designs, and character models, promising to significantly reduce development time and costs. As AI advances, it opens up new possibilities for creating immersive AI gaming experiences that were once impossible.
Smart AI NPCs
Non-player characters (NPCs) are crucial for driving the storyline in many games. Traditionally, NPCs have been scripted with limited dialogue and actions, often leading to a reputation for being "dumb." However, AI is changing this narrative.
At Convai, we are pushing the boundaries of AI-powered NPCs, enabling them to possess conversational AI capabilities, scene perception, and action. AI-powered NPCs create more engaging and dynamic interactions between players and virtual characters, particularly in simulation games and social virtual worlds. Convai gained significant traction after Jensen Huang demo-ed their product at GTC earlier this year.
While it can be fun for users to chat with AI NPCs, we hear from game developers the question: How would smart NPCs change the gameplay? The answer depends on the game genre:
- For PvP games like League of Legends or Counter-Strike, NPCs aren’t part of the core gameplay, so smart NPCs are a low priority.
- Simulation games like The Sims and social virtual worlds like Second Life have no/non-linear storylines, and smart NPCs would be the most exciting for this genre.
- For narrative-driven games, NPCs are just tools to advance the storyline. When playing these games, a player might stop to chat with an NPC for a few minutes for fun, but it’s more of a nice-to-have than a must-have. People don’t play games so that they can talk to NPCs. Convai’s Narrative Design feature allows developers to tell these stories with smart NPCs.
Convai powered AI NPCs add another dimension to the traditional smart NPCs. Convai’s AI NPCs can perceive their surroundings and are also able to perform different types of actions and interact with their environments.
AI-Generated Game Assets: Weapons, Skins, Maps, and AI Characters
Creating assets for games like Call of Duty, Skyrim, and Grand Theft Auto is time-consuming and expensive. For example, The Elder Scrolls V: Skyrim (2011) took a team of 100 people 5 years and $100M to make, with an estimated 10,000-20,000 assets, including 600 apparel items, 300 weapons, and 50 plants.
AI-powered asset generation promises to significantly reduce the time and money needed to develop a game. AAA game development studios like Blizzard and EA already use AI to generate concept art and digital sketches for game levels and challenges. Blizzard, for example, trains image generation models for each game (World of Warcraft, Diablo, and Overwatch) to generate concept art in the game's style.
EA uses AI to generate quick digital sketches for game levels and challenges. They claim that these designs used to take weeks to draw up by hand and can now be produced in hours.
As games move from 2D to 3D, AI is also used for photogrammetry, generating 3D models from 2D images. Startups like Alpha3D, Luma AI, and Kaedim work on this technology to create realistic 3D landscapes and characters.
However, while AI-generated assets look impressive in short demos, they may not be sufficient for extended gameplay. Companies use AI for smaller tasks, such as customizing assets or generating background noise and chatter for NPCs.
For example, Roblox created a tool that allows developers to use text prompts to customize an asset, e.g., changing a car’s color or turning on its headlights. Ubisoft has Ghostwriter, which uses AI to generate background noises and chatter for NPCs. GenAI is also great for adapting your assets to different seasons.
UGC and AI Toolings
User-generated content (UGC) has been a big factor in gaming, with players customizing the games they love through mods. Mods have proven influential, with some, like Counter-Strike, becoming massively popular standalone games.
Recognizing the importance of mods in a game's revenue, companies have introduced tooling to make it easier for users to create content and get paid for their creations. Fortnite and Roblox are two of the biggest creator economies in gaming, with Fortnite giving creators 40% of revenue from all of Fortnite and Roblox paying developers 25% of revenue from their creations. The success of digital fashion in these games is a testament to the thriving metaverse economy.
Convai is also exploring revenue-sharing opportunities for developers with our Convai Connect feature. This feature incentivizes developers to build more immersive experiences without worrying about the potential costs of running a popular game.
As AI further develops, many in the gaming industry are excited about putting generative AI tools in users' hands, which will significantly boost both the quality and quantity of UGC. Roblox has already introduced generative AI tooling for its creators, and it's likely that other games will follow suit.
Using AI for Video Game Analytics and Game Personalization
Developing a new game is risky, as getting a new title to top the charts is as challenging as having a book become a New York Times bestseller. Once a promising game attracts players, keeping them engaged for as long as possible is crucial, as user acquisition is expensive. Churn prediction is a complex task. There are many reasons why a player might stop playing, and it's difficult to determine these factors without direct communication with churned players.
Game personalization addresses the reasons for churn that the game itself can influence. For example, if a player churns because the game is getting old, adding new levels and quests might help retain them. For social players, introducing competitions or multiplayer challenges could be effective.
This has led to the rise of LiveOps, a process of continuously delivering new features, content, and in-game events to improve player experience. Games like Candy Crush and Clash of Clans, which have some of the highest retention rates in the industry, utilize real-time game personalization to adapt to each player's preferences and behaviors.
Improving Productivity Using AI
The applications of AI gaming are extensive, from AI-powered testing tools that automate the tedious process of identifying bugs to AI-driven localization services that make games accessible to a global audience.
Studios use AI to provide personalized customer support so players receive timely assistance tailored to their needs. Generative AI tools also provide copilot experience for game developers by assisting with coding, writing compelling game descriptions, and creating initial drafts of game assets.
As the AI gaming industry advances, it is poised to improve how games are developed, tested, and delivered to players worldwide.
AI Testing Products
Game testing is notoriously demanding, particularly in complex environments like open-world games. AI improves this aspect of game development, where tools like EA’s reinforcement learning agents drastically reduce the manual labor involved. For example, Battlefield V's AI testing scenario required automation for testing over 600 features, potentially saving around 300 work years if done manually.
Convai includes the Testing Framework feature, which uses AI to streamline testing processes for gaming platforms. The Testing Framework bridges the gap between the dynamic nature of AI-driven conversational agents and the need for reliable, predictable interactions. We believe that by offering this capability, we can significantly enhance the confidence and satisfaction of our users, ensuring their AI characters behave as intended, regardless of the complexities involved.
The steps involved in testing your AI Character:
- Modify the Character as needed.
- You can choose a subset or all test cases on the Testing Tab and run them with a single click.
- This will initiate a rerun of all selected test cases using the latest Character modification.
- The system generates and displays the new outputs upon completion, providing immediate insight into the testing outcomes.
Curious about Convai’s Testing Framework? Learn more about this feature in this blog.
Localization Using AI
Localizing a game for worldwide audiences requires adapting dialogues, characters, and assets to suit different cultures. AI offers powerful tools for this process. AI can translate vast amounts of text and synthesize convincing voice performances in multiple languages. Other aspects of localization, such as character design and cultural references, can also be refined using AI techniques.
AI makes localization faster and more cost-effective so that game developers can reach broader audiences with impactful experiences.
Convai provides multi-lingual support for game developers to integrate multiple languages, voices, and accents in virtual characters. You can select up to 4 languages for multi-lingual characters:
AI for Customer Support
AI gaming companies provide real-time, interactive help desks that improve player support and retention by integrating LLMs like ChatGPT into documentation. Players can ask questions directly about gameplay mechanics or technical issues, leading to faster resolution.
AI-powered customer support reduces the burden on human support teams while improving the player experience through readily available information.
Generative AI for Copilot Game Development
Generative AI is assisting developers in various creative processes. Tools like GitHub Copilot for code optimization, ChatGPT for crafting compelling game descriptions, and visual AI tools like Midjourney and Stable Diffusion for asset creation revolutionize game production.
These AI copilots accelerate development timelines and enhance creative outputs, demonstrating AI’s invaluable role in modern game development and aligning with the latest AI gaming trends.
Conclusion
AI is rapidly transforming the gaming world, from the content creation process to the player experience itself. As AI-powered tools become more accessible and sophisticated, we can expect to see even more groundbreaking innovations in game development and rapid developments in AI gaming trends.
Smart AI NPCs with rich personalities could revolutionize narrative-driven games, while AI-generated content could democratize the creation of expansive, customizable worlds.
Game studios are carefully integrating AI analytics to improve player personalization and retention. This intersection of AI and gaming trends promises an exciting future where AI acts as a creative partner for developers and enhances fun and immersion for players.
If you're eager to explore these topics further, check out our blog posts on Convai NPC Builder and Long-Term Memory.