AI, Automation and ML Solutions: 2022 Industry Results
Ahead of this year’s Global Game Quality Forum, we highlight the greatest challenges with implementing artificial intelligence (AI), automation and machine learning (ML) solutions in the gaming industry, and provide key insights from David King, Director of Technology at Electronic Arts (EA) on the results of our 2022 industry report.
What is your greatest challenge with implementing AI, automation and ML solutions?
Unexpectedly, around one-third of respondents chose 'Other' and provided a wide range of different challenges with implementing AI, automation and ML solutions. This suggests that the gaming industry is just beginning to explore these technologies. 'Knowing which processes are best suited for AI, automation and ML solutions', came in close second. Around a quarter of respondents felt that 'Knowing which technologies to use' and 'Alignment of strategy with business goals' were the greatest challenges, and less than 10% felt that 'Demonstrating ROI' was the greatest challenge with implementing these solutions. Therefore, it is extremely important for the gaming industry to invest in these solutions, and clearly define the steps required to best implement AI, automation and ML solutions, to overcome these challenges.
Insights from David King
The gaming industry has some unique challenges when it comes to implementing classical automated testing, AI, and ML solutions into their operations
"In the non-games software industry, testing is a more or less solved science. There are many different methods for developing and testing in an automated fashion and they continue to evolve, but the core principles are sound, well documented and researched. Games tend to be more complicated to effectively test, given that a lot of the behaviour is driven by data and content, as opposed to the underlying code. This precludes the usage of lower-level testing practices such as unit or integration tests. AI and ML approaches also carry a large amount of complexity given the worlds in which they are expected to operate. Understanding story context and navigating a 3D world is potentially a much more complex endeavour than image recognition. We can see this in the data on the challenges with implementing these solutions, that the results centre around how to best implement these solutions in each respondents’ individual projects. This is a challenge given that many ‘off the shelf’ solutions do not cater for the gaming industry challenges, particularly when it comes to multiplayer/multi-device testing."
Acquiring and growing expert skillsets will be key to successfully deploying effective automated testing, AI and ML solutions
"Automated testing, AI and ML are all fields which require lots of specialised knowledge, tools and training. There are roles in which people specialise in all of these in many industries and companies, and there are multi-year university courses focusing on just these areas. Getting this knowledge into the gaming industry, as well as retaining and growing this knowledge is going to be crucial for success. You can see this in the data, with respondents stating that their greatest challenges are ‘knowing which processes are best suited for AI, automation and ML solutions’, ‘knowing which technologies to use’, and ‘specialised training or courses for our developers’, highlighting the need for the industry to upskill in these areas."
Want to learn more about how you can implement AI, automation and ML solutions into your operations?
Join us in Amsterdam, to discover how you can increase efficiency, reduce costs and successfully implement AI, automation and ML solutions into your operations, at the Game Quality Forum 2022.
