DEV Community

Cover image for Generative AI in the world of video games
GIGO Dev
GIGO Dev

Posted on • Originally published at gigo.dev

Generative AI in the world of video games

Recently, I came across this interesting article on a couple monumental breakthroughs in the world of AI research.

There is a unique model that can create a playable 2D game from just a text description, a sketch, or a photo. Genie, revealed by Google DeepMind, is unique not only for what it can do, but how it was trained.

The model is able to learn controls from being trained on video footage. This is a big milestone in the world of generative AI.

This development got me to think about the video game industry a bit. With the constant emergence and evolution of generative AI, what does that look like in the world of gaming?

You may have heard of different game studios using AI in various ways and of course the response will vary between support and disapproval. Let’s dive into how AI is impacting the industry today and possible future implications.

Using Generative AI in Development

In general, the use of AI in video games is not anything new. In fact, its been around for decades. Over recent years, however, with the increase of AI research and development, tools such as generative AI can offer so much more to the industry.

Video game companies are beginning to explore generative AI’s potential in both development and gameplay. Rec Room’s Seattle-based creators introduced Fractura, a room demonstrating generative AI’s capability in content creation.

They used multiple AI tools, including ChatGPT, Midjourney, DALL-E, CSM, Shap-E, and Skybox, to develop Fractura’s environment and backstory. Despite industry challenges and job losses, these advancements suggest a bright future for 3D generative AI in gaming GeekWire.

The use of generative AI could streamline development efforts and be used as a tool to enhance the development cycle. Tasks such as playtesting, level design, and even NPC behavior could use generative AI. This would allow developers to focus on higher-level creativity in the game.

Interactive and Repository Gameplay with AI

With AI on a track to rapidly advance and expand, the use of generative AI in video games becomes more and more of a possibility. Imagine games with open worlds that react more and more with the players playstyle.

NPC interactions could feel more real and fluid. The world could have consequences to player actions, even outside major decisions within the narrative. It could also be used to adjust the gameplay in order to align with the players skillset.

Instead of setting a difficulty, enemies and tasks will become harder or easier based on how the player handles these situations. The possibilities are endless and could impact the future of gaming in the future (LinkedIn). But what are some drawbacks of using generative AI?

Let’s take the use of procedural generation in No Man’s Sky. Hello Games, the developers of the game, use procedural generation to create a near-infinite amount of diverse planets and universes. Each world is generated with its own unique features, environments, details, and more.

As someone who plays the game from time to time, it truly is impressive how vast the game’s world is. It feels cool to travel to a random universe and I am the first one to discover it. You will even come across new solar systems and their contents already discovered by other players, the same way they will see the things you discovered if they come across it.

To some, this could look similar to generative AI, but the procedural generation still allows some developer control, even at this scale. With generative AI, something of this scale could get out of hand. You may see levels that are unplayable, unintended deviation from the narrative, and even things that just don’t make sense in that particular setting (CNN).

Generative AI does not have to only be involved in gameplay or anything the player interacts with. There are tools that can be used to enhance the game engine. Studios can make the animations and graphics smoother with algorithms driven by generative AI tools (NVIDIA)(CIO).

What's Next?

As with anything where AI is involved, there is always the fear of job loss or even less of a human touch in design. In the video game world, however, it is so complex that AI wouldn’t necessarily replace individual jobs, but be used as a tool to enhance them.

Even if generative AI gets to a point where it is being used to create personal experiences in games, there will need to be teams of developers that know how to harness the AI in development.

With the everyday advancements in the world of AI, the use in video games are near endless. Whether you think it’s a good thing or not, it will become a reality that more AI tools will be used in the development of video games much like it already has for years.

The emergence of more advanced AI is nothing to be concerned with, but something to embrace. Industry’s will adjust and AI will become more of a tool than a replacement.

GIGO Discord

GIGO Twitter

GIGO Reddit

GIGO GitHub

Find this article on Medium
Find this article on Medium

Sources:

GeekWire — https://www.geekwire.com/2023/how-generative-ai-could-change-the-way-video-games-are-developed-tested-and-played/

Google Cloud — https://cloud.google.com/blog/products/gaming/generative-ai-in-the-games-industry

Andreessen Horowitz — https://a16z.com/the-generative-ai-revolution-in-games/

CIO — https://www.cio.com/article/1294946/game-on-the-evolution-of-gaming-through-generative-ai-innovation.html

NVIDIA — https://blogs.nvidia.com/blog/ces-2024/

IxieGaming — https://www.ixiegaming.com/blog/how-genai-changing-the-gaming-landscape/

LinkedIn — https://www.linkedin.com/pulse/how-generative-ai-transforming-world-gaming-giovanni-sisinna-kmqmf

CNN — https://www.cnn.com/world/generative-ai-video-games-spc-intl-hnk/index.html

Top comments (0)