Artificial intelligence has been a growing resource for video games for years now. Most video games — whether they’re racing games, shooting games, or strategy games — have various elements that are controlled by AI, such as the enemy bots or neutral characters. Even the ambiguous characters that don’t seem to be doing much are programmed to add more depth to the game, and to give you clues about what your next steps should be.
However, there are also many other ways that AI and game development are growing through each other. Although AI continues to be used to bring video games to life, video games are now being trained to study their own patterns so as to improve their own algorithms, which is just one of many ways that AI is becoming more advanced.
Trending AI Articles:
1. Deep Learning Book Notes, Chapter 1
2. Deep Learning Book Notes, Chapter 2
3. Machines Demonstrate Self-Awareness
4. Visual Music & Machine Learning Workshop for Kids
AI in Video Game Programming
In any video game environment, AI can be programmed to enhance a player’s experience. Machine learning runs on the stores of data available to it, and uses this information to create a world where characters can live and conduct basic actions. All of the basic information collected through machine learning can then be used to create game environments and characters that appear more realistic and natural.
However, to do this, AI algorithms must be provided with an abundance of information to have the proper reactions to specific stimuli. The enormous amount of data required to successfully train AI algorithms isn’t available to just anyone, and is why machine learning hasn’t yet been adopted in every industry, although its uses are endless.
The video game industry, however, is one of the technology niches where designers are expected to dedicate a lot of resources to investigate what AI is or is not capable of accomplishing. For video game creators who have access to bounds of technology, and have many uses for automation within storylines, the only barriers to fully using this advancing technology is the eventual limit of money and time.
Theories and Algorithms
Practical AI for video game development involves applying algorithms, such as the Minimax algorithm, into machine learning in order to provide the technology with all of the information it needs to outsmart a human. After applying game theories and formula information into AI technology, scientists found that they could program a computer with enough information to beat professionals at no-limit Texas hold’em.
According to Author and Engineer, George Epstein, “The AI researchers used game theory — the mathematics of strategic decision making — to find the best strategy for each hand, while faced with a variety of uncertainties. Because the possibilities are so vast, this usually involves making appropriate approximations — quite a task!”
AI is able to do this because the information it holds brings each decision down to a science. All of the information used by machine learning came from humans, but with enough information from enough resources, the machine has more strategic knowledge than is possible for a human to have in mind at all times.
Current Video Games Helping AI
The virtual worlds inside video games provide machine learning with a flexible environment for quick changes and easy customization. Because of this, video games are being used to train AI software to understand a large variety of situations.
Grand Theft Auto
One of these situations involves the use of the video game Grand Theft Auto to help the development of autonomous vehicles by training them to recognize stop signs, including those that were partially obscured by shadows, dirt or weather. Although the software needed to be adjusted slightly for a computer to play it, the virtual world inside the game provided the AI algorithm with a semi-realistic application for it to learn from.
Another video game that has been used to train machine learning algorithms is the action-adventure stealth game Assassin’s Creed. This series of video games uses high-quality computer-generated imagery (CGI) throughout its storyline that could be mistaken as a film.
Computer scientist Adrien Gaidon, of the Xerox Research Center Europe, used Unity, which is a popular game development engine for 3D video games to develop scenes that would help train AI algorithms. By using advanced CGI environments, as well as real-life scenes, the scientists were able to compare the effectiveness between the two in training AI algorithms.
The sandbox video game Minecraft allows users to build with a variety of different blocks in a 3D procedurally generated world, and has been used by Project Malmo to conduct AI experiments that support important research into machine learning. Although Minecraft’s block-designed world doesn’t seem like it would be the best option for teaching AI how the world really looks, the submersion that players face in the Minecraft environment can be the same for AI technology.
The world generated in the game creates an environment where AI can learn from its surroundings and develop an internal representation of the space, and can help researchers understand the learning process and machine learning perspective of the world.
Data Analysis and AI
As the AI industry grows and continues to expand, not only in niche massive-multiplayer online role-playing video games, but across the technology fields of all industries, software developers must be prepared to work within machine learning. With the excess technology available to us today, business decisions are becoming more evidence-based, requiring a data-driven approach to business operations. One of the most effective ways to use the highly sought after big data people are so eagerly collecting is to create machine learning software that can be implemented to organize data for easy use.
Consumers now have so much access to information and are so technology savvy that game developers are required to step up their own already profound knowledge. As big data becomes more pervasive across industries, game developers and other creators are needing a more analytical approach to information. This is a common aspect of video game development.
AI will continue to have a huge impact on the video game industry and game development. As information becomes more accessible and simplified to the average game and software developer, we’re likely to see a huge shift to more advanced visuals and characters that are able to create their own storylines. The advancing development of AI in video games is having impact outside of the gaming industry, a trend that is likely to continue.