Technology

How People Can Use AI To Create Their Perfect Game

Share this article

Share this article

Technology

How People Can Use AI To Create Their Perfect Game

Share this article

Artificial Intelligence (AI) is revolutionising literally everything - and there are those who are for it - those against it… and those who are sitting on the fence and seeing how everything pans out. Whilst there are obviously some concerns about what it means for people… job replacements - and the fact that we don't even know if we can believe what we read or see anymore, there are definitely some huge advantages to be had.

In the gaming landscape, it is starting to be widely used. Not only to help create more immersive experiences when playing online - with AI being used to personalise gaming preferences and even in the creation of sign-up bonuses offered by gambling sites - but it is also starting to be used to help create games.

There is a relationship between AI and gaming that has an incredibly interesting history, from earlier attempts at creating competitors at Chess to more contemporary endeavours with more complex games such as Go.

However, with more recent developments, AI's role in the gaming industry is reaching new heights - allowing gamers to become creators of their own unique gaming experiences.

In the Earliest Stages

When AI was at its very beginnings - and just being introduced to the world, it was games that served as benchmarks for intelligence. One of the very first testing grounds for AI researchers looking into how it could potentially work was Chess, along with games including Pong and Super Mario Bros. These became arenas for mastering complex decision-making.

Then, as AI evolved into machine learning, video games started to emerge as potential testing grounds. Essentially, games offer a fully observable environment with clear reward systems - which is ideal for training machine learning algorithms.

The Technology Improves

Matthew Guzdial, an AI researcher and assistant professor at the University of Alberta, is one of the names right at the forefront of the use of AI in game creation. His work essentially involved developing algorithms that could collaborate with human partners in order to generate entirely new video games.

The process basically involves training an AI machine to learn and understand the rules of existing games, such as Super Mario Bros., for example, by watching and analysing hours of gameplay.

The AI, having observed these videos, then attempts to establish the rules - and refines its understanding of the game by watching and re-watching the videos. Once the rules are fully established, algorithms merge this knowledge into a representation known as the 'game graph'. This web of information will then become the foundation for reproducing and creating new elements within the game.

The AI will then utilise this game graph to design, combine, and reproduce what it has learned - resulting in the creation of entirely new games. So, for example, the AI could potentially blend the platforming styles of a few different games to create a whole new experience. Most importantly, Guzdial ensures that his work is fair and ethical…, avoiding any copyright issues that could be associated with game creation.

The objective of this AI-assisted game creation isn't necessarily to replace game developers but to make the process more democratic. Guzdial states that he would like to lower the barrier to entry for game creation, allowing gamers to be able to build their own levels on games without needing to study coding for years. This is similar to tools available in games like Nintendo's Super Mario Maker, where players are able to design their own levels and share them with the gaming community.

Looking ahead, Guzdial aims to develop a tool that enables gamers to instantly create entirely new games by defining their desired look and feel. The AI would then use this input to design a complete game, complete with rules and levels. The key innovation here is that the AI system requires minimal input, only two frames worth of data defined by the player. The iterative process involves the AI learning rules, making predictions, receiving feedback, and adjusting accordingly, resulting in a brand-new gaming experience.

The Future of AI Gaming

Despite the incredible potential of AI in game creation, it's fair to say that a consumer-ready version of this technology is still quite a long way away. At the moment, the gaming industry could well be a little wary about widespread adoption - worrying that AI could potentially overshadow the creative contributions of human game designers.

Guzdial believes that it will be independent developers who will be the pioneers - as they experiment with these technologies to create more unique gaming experiences.

Ultimately, the combination of AI and gaming has already gone way beyond what we could have expected - now allowing those even with no coding or technical prowess to create their own virtual worlds. As AI keeps getting better and better, it looks set to make game development even more accessible and enjoyable to a much wider audience - and it's exciting to see where it could possibly go.

Get news to your inbox
Trending articles on News

How People Can Use AI To Create Their Perfect Game

Share this article