Algorithms that can beat grandmasters and NPCs that seem too lifelike to be fake are just a few examples of AI’s central role in gaming today. AI has always been a part of the modern gaming industry, even in retro titles like Pac-Man. However, exciting developments are on the horizon and AI could soon revolutionize the gaming industry.
The illusion of intelligence
Developers design today’s game AIs to create an illusion of intelligence. They seem smarter than they really are by giving the player intelligent cues that instill the feeling of realism and randomness. A viral example of this is a now-famous clip from the Western game Red Dead Redemption 2 where a player randomly fires their gun into the air and shoots down a bird — an incredibly rare, one-in-a-billion event.
Game worlds like this use background AI to create an illusion of randomness in a vast, lifelike game world. NPCs may seem as intelligent as real creatures. AI-procedural world generation can lead to immensely complex game worlds, such as those found in the ASCII-based game Dwarf Fortress.
In this resource management strategy game, the AI can generate worlds so vast they are known to take up to 10 or 15 minutes to develop. Each one is populated by hundreds of NPC characters, each with their own AI-operated behavior. In fact, Dwarf Fortress was considered the future of video games early in its development.
AI game bots have become highly sophisticated, as well. Algorithms have been able to beat grandmasters at chess since the late 1990s. AI can even take on some of today’s most complex games — like DOTA 2 and StarCraft — with algorithms giving even the best players a run for their money.
Despite how advanced these uses of AI seem, they are actually quite restricted, as game developers have an interesting relationship with these algorithms. Rudimentary AI is vital for creating a fun experience for players, with realistic NPCs and exciting worlds. However, giving the AI too much freedom in-game can lead to an out-of-control level of randomness that makes games unplayable. So, while AI may remain an illusion of intelligence in-game, it will still have a pivotal role behind the scenes.
Next-gen AI game development
AI will be improving the gaming experience in the near future, but it will likely be happening behind the scenes. A game developed entirely by an AI with completely generative play would probably be unplayable — a mess of unpredictable gaming with no cohesive world or storyline. However, AI can be extremely useful as a tool to help developers create games more efficiently and realistically.
Seeds of this next-gen approach to game design are already appearing in the industry. The top game development degree programs are prioritizing cutting-edge technologies, particularly AI models. Tomorrow’s game developers will have to create realistic worlds while balancing the computing demands of VR and 4K gaming, which will require the help of AI. Art generation algorithms — such as samples of the DALL-E 2’s eerily remarkable artwork — are going viral on social media already.
AI-generated game elements
Soon, AI will assist game developers in creating realistic worlds that don’t cost as much as a Hollywood blockbuster. For instance, it can take dozens of hours for a game artist to create a single building or vehicle, costing thousands of dollars.
AI art generation can cut down this time. Developers can train the algorithm to reproduce a particular art style and have the AI generate assets matching that. The artists then choose and refine the best ones, developing a game faster and for less money.
NVIDIA is pioneering research in AI game development like this. They have developed generative adversarial networks — or GANs — that can do some incredible things. For instance, one GAN can recreate Pac-Man with no underlying game engine to give it the rules. Another can automate the animation of facial expressions. Others can create 360° virtual backgrounds, generate music and lyrics, and recreate a real person’s facial expressions.
Developers can use machine learning to create algorithms that understand movement and automate realistic animations. AI can replicate human-created art styles, generate unique faces for NPCs and automate the complex computing needed for ray tracing — a key element of hyper-realistic games. It can generate vast, unique worlds for the player while maintaining predictable game mechanics and graphics stylization.
Balancing art and AI
Completely AI-generated games would likely be too unpredictable and random to be fun for players, but that doesn’t mean AI isn’t instrumental to the gaming industry. Game developers could soon be using AI to create vast, detailed worlds for players to explore without taking years to complete and costing millions of dollars. This will make games more accessible, immersive and realistic — a win for everyone.
Devin Partida is the Editor-in-Chief at ReHack Magazine, where she covers topics related to data, cybersecurity, tech investments, and more.
TechNode Global INSIDER publishes contributions relevant to entrepreneurship and innovation. You may submit your own original or published contributions subject to editorial discretion.
3 vital developments supplementing digital banking industry growth