8 Modern Examples of Artificial Intelligence in Gaming
AI has been a buzzword in many industries over the last eighteen months. And when it comes to artificial intelligence in gaming, though it has been a constant companion for decades, the last two years have seen a rapid advancement in the use of AI, machine learning, and other tools. So let’s take a look at some recent examples of how AI is helping the video game industry evolve with the times with new tools, methods, and techniques.
Activision Blizzard Using AI-Generated Music In Their Games
Last November, ODSC reported how the studio giant, Activision Blizzard, patented a new AI-powered system aimed at immersing a player deeper into the game. Instead of just generating music, this machine learning model would generate music that was tailored to how the player was playing the game using algorithms. The way it worked is that it would take data from the player’s actions, movements, choices in the story/quests, etc. to study patterns and behavior. Then, it would take all of this, and generate music specific to the playstyle.
Skyrim Mod Powered by ChatGPT Gives NPCs Memories
It’s not just the largest gaming companies that are looking to AI as a means to enhance the player experience. A clever Skyrim player created a new mod this past spring that allowed for in-game NPCs to hold memories, and even generate conversations with the player. All of this was thanks to using ChatGPT. This may be the first major use of AI within the mod community, but it’s also unsurprising that the game was Elder Scrolls Skyrim, as it has been a testing ground for mod makers for close to a decade now.
But how does the mod work? Well, it seems that the modder uses a combination of ChatGPT, xVASynth, and Whisper. This allows for a memory system for each NPC that remembers past conversations with the player. ChatGPT is given the basic background of the character and asked to roleplay as them, while Whisper converts the player’s voice to text, which is then passed to the NPC to respond appropriately using ChatGPT. This system allows NPCs to recall previous encounters with the player, making conversations more realistic.
Ubisoft Introduces “Ghostwriter” AI-Powered Video Game Dialogue Generator.
Games aren’t getting smaller, instead, studios are pushing the scale and limits with each new release. This means that thousands of lines of dialogue have to be written for specific characters. This can be time-consuming, expensive, and at times, slows down the development process by pulling writers from other tasks, but this is where Ghostwriter from Ubisoft comes in. Back in March, the studio introduced this new AI in a bid to control the resource costs of creating massive amounts of dialogues for open-world games. That chatter heard in the background is a key element of immersing a player in the environment. Think of games such as Cyberpunk 2077, The Wicher 3: Wild Hunt, and of course Skyrim.
So the way Ghostwriter works is that narrative writers input characters and the type of interaction they’re looking to create within the program. Then, using machine learning, product variations of dialogues are reviewed by the writers. This speeds up the development process as it makes it much easier to create tens of thousands of pieces of dialogue.
AI-Generated Art Used in Rick and Morty Creator’s New Game: High on Life
Another interesting effect of AI is of course through art. Though artists have pushed back against the use of AI to create art in 2022, it didn’t stop studios from utilizing the tool to generate art assets for their games, and High on Life is an example. In the game, MidJourney was used to create art, but according to Justin Roiland, creator of Rick and Morty, these pieces were just the “finishing touches” of an already created game world. Though that may be the case, for smaller studios, AI helps to reduce the time and resource costs associated with art assets.
Matt Booty of Xbox Game Studios Wants AI-Powered QA in Gaming
Late last summer, the Head of Xbox Game Studio, Matt Botty commented how he was hoping that with AI and machine learning, they would see the technology help with quality assurance testing. His reasoning has to do with human QA testing being unable to keep up with the pace of content creation, creating a lag that could either hold a project hostage or see a game released with bugs. In his view, AI-powered testing bots could be “transformational” for the gaming industry due to how the bots could address these problems.
As of right now, there aren’t major leaps when it comes to QA bots, but it’s likely that at some point they will come out in one form or another. At first, they’ll support existing human QA testers.
NVIDIA Plans on Using AI to Transform NPCs in Gaming
During their annual computer expo, Computex, in Taipei, Jensen Huang, Nvidia’s CEO and co-founder, delivered an insightful talk about AI and hardware. But during this talk, he opened up about ACE or Avatar Cloud Engine, a new custom AI service that can be deployed by game developers for the purpose of supercharging their in-game NPCs.
The goal of ACE is to provide real-time conversations and animated facial expressions. It also allows NPCs to generate backstories, personalities, and goals within the game world. This allows the player to interact with an NPC. Though this might sound familiar, there is one aspect of this that is quite unique, as part of ACE’s cards is the ability to give NPCs automatic speech recognition. This means they will be able to respond to a player’s speech patterns.
Voice Actors and Generated Voice Dialogue.
Similar to concerns about QA testers being replaced by machine learning bots, voice actors face similar challenges due to generative AI models. In a report from Fagenwasanni.com, voice actors found their voices have been used to create new mods to create new dialogue from modders without their consent. As Tim Friedlander, the founder, and president of the National Association of Voice Actors (NAVA) spoke on the issue of AI and how it threatens voice actors. In his view, voice actors are in the same position as writers, where the technology will be able to learn from their work and generate similar content. His worry of course, in the long term, is that AI could be used to push human labor out of the equation.
To combat this, NAVA has allowed actors to include situations in their contracts where their voice recordings and motion capture performances can’t be used to create new content via AI. Though it’s a first step, this doesn’t stop independent actors, such as modders, from utilizing the data available to create new content.
Generative AI May Help You Design Your New Game Character.
Anyone who has played an RPG or other similar-style game knows that character creation menus can be fun and frustrating. They provide a window for people to create custom avatars to adventure in fantastical worlds. Now, thanks to a new generative AI tool, players will have even more customization options. In a report from May, Blizzard was actively looking to use generative AI to improve gameplay experiences by allowing players greater freedom in character creation.
In an internal email to fellow Blizzard employees, he called the new tool Blizard Diffusion, a nod at Stable Diffusion. Of the tool, he said in part, “Prepare to be amazed…We are on the brink of a major evolution in how we build and manage our games.” Though there is little information on Blizzard Diffusion, the legendary studio has pushed for greater use of AI over the years as a means of improving player experience.
Conclusion
It’s clear that the gaming industry has a lot going on when it comes to generative AI. From dialogue, music, immersion, smart NPCs, and more, studios are looking to artificial intelligence in gaming as a means to be more creative and provide a better experience for players. If you want to keep up with the latest in generative AI, then you won’t want to miss ODSC West this October 30th to November 2nd. There, you’ll engage with the experts in the field, and see for yourself what’s on the AI horizon.
Originally posted on OpenDataScience.com
Read more data science articles on OpenDataScience.com, including tutorials and guides from beginner to advanced levels! Subscribe to our weekly newsletter here and receive the latest news every Thursday. You can also get data science training on-demand wherever you are with our Ai+ Training platform. Subscribe to our fast-growing Medium Publication too, the ODSC Journal, and inquire about becoming a writer.
