Have you ever beaten a video game?
Props to you if you have.
But have you ever watched the end credits after completing the game?
So. Many. People.
All of these game creators from all walks of life — from programming to art design, narrative design, level design, QA, marketing, legal…
Video games generate more revenue every year than movies, TV, and radio put together, and the gaming industry is expected to generate a staggering $521 billion by 2027.
Standards are rising ever higher for bigger, better games — particularly in the AAA game market, where one wrong move lands you in very hot water with the player and critic reviews.
As demand and expectations climb, game developers need more help than ever before.
So how can developers expect to keep up? Artificial intelligence in video games could provide the answer.
AI has already swept across almost every digital industry as we know it — from cybersecurity to healthcare. As a result, many gaming studios are now looking to AI to assist them with their creative efforts..
Here, we look at how AI is helping game developers to create games faster, examples of AI in gaming, and some specific AI tools for game development.
What is AI in video game development?
AI is still a relative newcomer to the world of video gaming, but one that’s set to change the game design and development landscape forever — when used correctly. Its main focus for gaming studios is to create better user experiences for players, through technical and creative means.
Video game developers use AI to complete certain functions both behind the scenes in the back-end, or on the big screen through the user interface.
For instance, many AI tools are now used for both dynamic engagement through in-game features such as NPCs (non-player characters), and behind-the-scenes development (like data gathering for personalized gaming experiences.)
By using AI in this way, game developers can unlock a whole new world of opportunities within their games without needing to waste time, energy, or team capacity.
How AI is used to accelerate game development
AI for game developers is constantly evolving, and gaming studios large and small are already utilizing AI in several ways:
Non-player characters (NPCs)
A game without NPCs always feels eerily empty. But sometimes that can be preferable to bad NPCs — whether they suffer from strange dialogue writing to poor voice acting.
Just remember that “Jill sandwich” scene in the first Resident Evil game…
Thanks to AI, NPCs can be given a whole new lease of life.
It’s also the most easily visible way that AI can take center stage when deployed in games.
It’s particularly seen in story-driven RPGs (role-playing games) that require hefty amounts of dialogue from the NPCs, either with the player character or each other through passing snippets of conversation.
As technology has improved and games have grown more complex, users expect the world in the game to react more organically to them.
In the past, NPCs had pre-written scripts and dialog trees. Now, more and more games are giving them the ability to react to players’ choices and interact in more “human” ways. This could be through adaptive dialogue, reacting to player actions, or having “daily schedules” that can shift, for instance.
Scene design
Scenes and game worlds are getting bigger and more intricate. As scene design in video games gets more complex and detailed, gaming engines have to render and build massive spaces in a short amount of time.
AR (augmented reality) and VR (virtual reality) games in particular require a huge amount of scene design to better incorporate realistic graphics into the game for immersive gameplay.
Using AI lets developers create more realistic scenes with better visuals and graphics by using machine learning tools (ML) like deep learning to find better backgrounds and generate scenes on the fly.
Procedural content generation
This is a more future-facing one, but it’s already being implemented in some cases.
For instance, the popular procedural space exploration game No Man’s Sky has 255 galaxies for players to discover. This means that to explore all the planets in these procedurally generated galaxies, a player would have to visit one planet per second for nearly 600 billion years.
This means that experiences are unique, but they also need to be built automatically.
Relying on AI for procedural content generation accomplishes a lot — creating unique experiences, increasing replayability, making worlds and games that can scale more easily, and making discovery more engaging.
AI tools for game developers
Now that we’ve discussed some of the best use cases for AI in game development, let’s take a look at some of the AI tools on the market that are being used in games right now.
Charisma.ai
“I used to be an adventurer like you. Then I took an arrow in the knee.”
If you got the reference — and you were sick of hearing it every time you traveled around Skyrim — you’ll be pumped to hear about Charisma.ai, a tool for game character development.
This tool auto-generates NPCs who can provide appropriate emotional responses, remember player interactions, and even use AI-generated voices to make each character unique.
It’s also already being used by several major gaming companies like Epic Games and Sony Interactive Entertainment (through PlayStation), as well as big names like Warner Bros., Dreamworks, and the BBC.
Scenario
Typically, we find that game developers fall into two camps.
You either really want to focus on the artistic aspect of your games…or you really, really don’t.
If you’re the latter, or someone that wants to be in the first camp but just doesn’t have the time, then Scenario will be a game changer.
Literally.
Scenario is an AI tool for proprietary art, which lets studios and developers build custom art generators and AI engines but without the looming concern over image ownership.
Although artificial intelligence in game design is still fairly new, Scenario lets you train models to create content and graphics that are similar to your pre-existing assets, so you can continuously create gaming artwork that you can fully call your own.
Leonardo.ai
Similar to Scenario, Leonardo.ai also creates in-game assets and graphics.
Its crowning glory, though, is its ability to quickly create multiple versions of assets to find the one that works best for your game’s creative brief.
Say you have a general idea for a game asset — let’s use the example of a tiny fairy companion — but you’re not quite sure what the character should look like.
Leonardo’s rapid ideation and iteration means you can not only generate new assets for the character art, but also keep regenerating them, tweaking as you go, until you’re happy with the end result.
If you fancy a less hands-on approach to AI game design, it also comes with pre-trained models that can auto-generate assets ready for production.
Which is very helpful if you’re staring down the barrel of an incoming crunch and need fast results.
Promethean AI
Where tools like Scenario and Leonardo focus on the artistic aspects of game development, Promethean AI does the rest of the hefty work involved in the back-end development.
As such, Promethean is like the mortar to the bricks — holding everything together to produce one finished piece.
It’s used by studios and developers to assist in non-creative tasks, from digital asset management to virtual world creation and design.
Overall, it makes the process easier and fills in any gaps in expertise or resources.
Latitude
Latitude, the creator of the AI video game AI Dungeon, a text-based adventure title, has now ventured into creating more generative AI tools for game devs.
Developers and narrative designers alike can use this tool, which acts like ChatGPT mixed with Twine for world-building.
It’s capable of creating immersive, unique worlds that feature significant possibilities and potential branches.
Perfect for weaving in juicy plot twists without needing to become an Oscar award-winning scriptwriter.
Modl.ai
QA and testing for video games can be a monumental effort at the best of times, with hundreds of testers trying everything to eke out any signs of glitches, bugs, and other unexpected performance issues.
Modl.ai is a custom bot generator that can be used for automating and testing different aspects of game development. Developers can create an unlimited number of unique users to test their games and find glitches, bugs, and other unexpected performance issues
Accelerate your game dev with AI
It’s clear that AI is already being adopted by game developers around the world — from small creative indie studios all the way to the lofty heights of AAA game companies. As more artificial intelligence tools develop over time and scale to the industry’s demand, we fully expect to see a new wave of AI-generated video games coming very soon.
