A gamer uses a computer powered by an Nvidia Corp. chip. at the Gamescon video game trade show in Cologne, Germany on Wednesday, August 23, 2023. Gamescon runs through Sunday, August 27. Photographer: Alex Kraus/Bloomberg via Getty Images
Bloomberg | Bloomberg | Getty Images
It’s not just human life that will be remade thanks to rapid advances in generative artificial intelligence. NPCs (non-playable characters), characters who populate generated worlds in video games but have until now had to operate largely on limited scripts – think of the store owner you walk into – are being tested as one early core aspects of the game where AI can enhance gameplay and immersivity. A recent partnership between Microsoft Xbox and Inworld AI is a great example.
Better dialogue is only the first step. “We create the technology that allows NPCs to evolve beyond predefined roles, adapt to player behavior, learn from interactions, and contribute to a living, breathing game world,” said Kylan Gibbs , Chief Product Officer and Co-Founder of Inworld AI. . “AI NPCs are not just a technological leap. They represent a paradigm shift for player engagement.”
This is also a big opportunity for gaming companies and game developers. Moving from scripted dialogue to dynamic player-driven narratives will increase immersion in a way that drives replayability, retention, and revenue.
The interplay between powerful chips and games has been part of Nvidia’s success story for years, but there’s now a clear sense in the gaming industry that it’s just starting to get to the point where AI will take off, after some initial uncertainty. .
“All developers are interested in the impact of artificial intelligence on the game development process,” John Spitzer, vice president of development and performance technologies at Nvidia, recently told CNBC, and cited feeding non-playable characters as a key test.
It has always been true that technological limitations and possibilities overdetermine the game worlds developers can create. According to Gibbs, the technology behind AI NPCs will become a catalyst for a new era of storytelling, creative expression and innovative gameplay. But much of what’s to come will be “games we still have to imagine,” he said.
Bing Gordon, an advisor to Inworld and former chief creative officer at Electronic Arts, said the biggest advances in video games in recent decades have come from improvements in visual fidelity and graphics . Gordon, who is now chief product officer at venture capital firm Kleiner Perkins and sits on the board of gaming company Take-Two Interactive, believes AI will remake the world of the game player and designer .
“AI will enable the creation of truly immersive worlds and sophisticated narratives that put players at the center of the fantasy,” Gordon said. “Additionally, AI that influences fundamental game mechanics has the potential to increase engagement and draw players deeper into your game.”
The first big opportunity for the AI generation could be game production. “That’s where we expect to see a major impact first,” said Anders Christofferson, a partner in Bain & Company’s media and entertainment practice.
In other professional tasks, such as creating presentations using software like PowerPoint and early drafts of speeches, Generation AI is already completing days of work in minutes. Initial storyboard design and NPC dialog creation are designed for the AI generation, which will save developers time to focus on the more immersive and creative parts of game creation, Christofferson said.
Create unpredictable worlds
A recent Bain study noted that AI is already taking on some tasks, including pre-production and planning of game content. Soon, it will play a larger role in the development of characters, dialogue and environments. According to Bain’s research, gaming executives expect AI to handle more than half of game development within five to 10 years. This may not result in lower production costs – blockbuster games can generate total development costs of a billion dollars – but AI will allow games to be delivered faster and with higher quality. improved.
Ultimately, the proliferation of the AI generation should enable the game development process to include the average gamer in content creation. That means more games will offer what Christofferson calls a “creation mode” allowing for an increase in user-generated content – Gibbs called it “player-driven narratives.”
The current shortage of human talent, a labor problem that exists in the software engineering field, is not something that AI is going to solve in the near term. But it could allow developers to spend more time on creative tasks and learn how to best use the new technology as they experiment. A recent CNBC study found that among the workforce, 72% of workers who use AI say it makes them more productive, which aligns with research conducted by Microsoft on the impact of its Copilot AI At work.
“GenAI is very young in the gaming space and the emerging landscape of players, services, etc. is very dynamic and evolving day by day,” Christofferson said. “As with any emerging technology, we expect a lot of learning to take place regarding GenAI over the coming years.”
Given the magnitude of change happening in video games, it may simply be too difficult to predict the scale of AI at the moment, says Julian Togelius, associate professor of computer science and in engineering at New York University. He summarized the current state of AI implementation as a “medium-sized deal.”
“In the game development process, generative AI is already used by many people. Programmers use Copilot and ChatGPT to help them write code, concept artists experiment with Stable Diffusion and Midjourney, etc.,” said Togelius. “There is also great interest in automated game testing and other forms of AI-augmented quality assurance,” he added.
The Microsoft and Inworld partnership will test two of the main implications of AI in the video game industry: design time and narrative generation assistance. If a game contains thousands of NPCs, having the AI generate individual stories for each of them can save considerable development time – and having the generative AI work while players interact with them NPCs could also improve gameplay.
The latter will be trickier to achieve, Togelius said. “I think it’s much harder to achieve, partly because of the well-known hallucination problems of LLMs, and partly because the games aren’t designed for that,” he said.
Hallucinations occur when large language models (LLMs) generate responses that deviate from context or rational meaning – they speak nonsensically but grammatically, about things that make no sense or have no meaning. relation to the given context. “Video games are designed for predictable, hand-crafted NPCs who don’t deviate from the storyline and start talking about things that don’t exist in the game world,” Togelius said.
Traditionally, NPCs behave in predictable ways, hand-written by a designer or design team. In fact, predictability is an essential element of the video game world and its design process. Open-ended games are exciting because of their sense of infinite possibility, but to work reliably they incorporate great control and predictability. Unpredictability in the gaming world is a new area and could pose a barrier to wider use of AI. Finding this balance will be key to moving AI forward.
“I think we’re going to see modern AI in more and more places in gaming and game development very soon,” Togelius said. “And we will need new designs that exploit the strengths and weaknesses of generative AI.”
Gn tech
