Mobilise CEO Kamal Krishna decodes AI’s impact on mobile gaming

In this insightful interview with Adgully, Kamal Krishna, CEO of Mobilise, delves into the diverse applications of artificial intelligence (AI) in the mobile gaming ecosystem. From the evolution of non-player characters (NPCs) in games to real-time decision-making algorithms, machine learning in game development, procedural content generation, and the delicate balance between AI-driven automation and player agency, Krishna provides a comprehensive overview of the current landscape and future trends. He discusses the transformative role of AI in shaping the gaming experience and the potential challenges and ethical considerations that lie ahead. 

What are the use cases of AI in the mobile gaming ecosystem, such as the role of AI in creating realistic non-player characters (NPCs) in games? 

In the past, accountants who did not fear calculators and adopted them instead, have made the idea so mainstream that we do not think twice about its ubiquity. 

One example in that light is the AI integration in GTA 6 development that has helped reduce, as well as automate a whole lot of menial tasks, which in turn has allowed game designers to keep their core focus on high-level creative work. This is incredibly important for gaming companies, who are otherwise constantly battling for inches within gamer minds, and being more creative helps stretch that. 

Gaming environments have improved as well, with NPCs (Non-playing characters) evolving to become more than just scripted. These are now extremely dynamic and engaging, in turn contributing to a much more immersive and interactive gaming experience. 

Lately, DeepMind has been rocking the AI world, creating systems that play games like Go, Chess, and StarCraft II even better than the pros. They’re not just doing it for the game’s glory; it is part of a bigger plan to make AI as smart as us in various tasks, including gaming. Think about it – AI has come a long way since Garry Kasparov took on IBM’s Deep Blue in that epic chess match. Nowadays, AI is shaking up the gaming scene, giving us NPCs that feel more real, evolving and learning as you play. 

How do AI algorithms enable real-time decision-making for NPCs in dynamic gaming environments? 

In recent times, Electronic Arts established a dedicated R&D division known as SEED (Search for Extraordinary Experiences Division), which focuses on exploring cutting-edge technologies and creative opportunities that AI can unlock for the future of AI games. We are living in an age where AI algorithms can adjust the difficulty within gameplays looking at a player’s performance – whether the person is a noob, or more experienced. If a player struggles at a particular level, such interventions via gameplay and NPCs ensure that the player remains engaged and doesn’t get frustrated. 

In other use cases, AI is helping define the states and transitions for NPC actions based on inputs and conditions; for example, Behaviour Trees are being used to define the logic and sequence of behaviours for NPCs; while neural networks are utilised for modeling the learning process and decision-making of AI-driven NPCs. 

How is machine learning used in game development? What challenges do developers face when implementing machine learning algorithms in games? 

In GTA 5, the game developers used ML algorithms to take not-so-great images and make them way clearer using fancy techniques with deep learning and neural networks. This approach allowed them to enhance the visual quality of the game, making it more photorealistic and immersive for players. They fed the computer with lots of fancy pictures and trained it to upgrade the blurry textures to clearer ones. This trick helped find patterns and improve the not-so-clear textures. FIFA, a well-known game, also does something similar with smart ML algorithms to control complexities. It examines all the team data and checks how players act. The algorithms even keep an eye on how players move around the field and their positions, making sure they act just like real players. 

Challenges around heavy computational power that’s required, and also the need for large amounts of data for ML are well established as being expensive and oftentimes insurmountable. There are also ethical, consent, and privacy-related issues on the one hand, and design related trade-offs that are made when ensuring ubiquity and uniformity of experience across device and player types. Game designers must navigate these issues responsibly. Ensuring player privacy, obtaining informed consent, and being transparent about data use. 

How can AI be employed for procedural content generation in game development? What are the advantages and potential drawbacks of using procedural generation in gaming? 

Most PCG methods have been developed because of the need for new storylines and to make the games replayable. PCG is the secret sauce for crafting diverse game levels, quests, and challenges that guarantee players a unique experience. Imagine having a game that understands your style and keeps getting more exciting with every move you make – that’s the power of AI in PCG. 

For example, in such cases, using predetermined parameters and rules, the algorithm can be used to design a certain kind of never-ending planet based on distance from the star or sun, presence of elements, etc. 

Challenges remain, though. Procedurally generated content may lack the hand-crafted detail and design that human developers can provide without careful design and curation. PCG also runs the risk of being repetitive or uninteresting for gamers in some cases. 

How should game designers balance AI-driven automation with maintaining player agency and control? 

Game designers must prioritise player agency, ensuring that AI-driven automation enhances, rather than replaces player control. By incorporating customisation features in games, companies can allow players to adjust the level of AI-driven assistance based on personal preferences. For example, Fortnite’s AI system learns from each player’s actions and preferences, and tailors the game for a personalised experience. AI-powered tools can also streamline game design and development process, allowing developers to focus more on crafting compelling narratives and innovative gameplay mechanics. 

What will be the future of AI and gaming? What will be the future trends? 

Cloud-based gaming is a no-brainer – it’s a standout area for AI in gaming. With this, gamers can play their favourite games from any device, anywhere, and without the need for high-end hardware. Prevalence of 5G is already bringing improvements across mobile e-sports games in terms of graphics, animations, and visual effects. As AI becomes more integral to gaming, ethical considerations around data privacy and algorithmic bias will continue to gain prominence. Also, AI can enhance Virtual Reality (VR) and Augmented Reality (AR) games, making them more immersive and interactive for the gamers. 

AI can also be used for player-experience modeling (PEM), data-mining, and real-time analytics, providing insights into player behaviour and preferences. Also, Natural Language Processing (NLP) analyses in-game chats to understand emotions and tweak the game accordingly. 

Despite the hurdles, data scientists have been aiming to infuse real and genuine emotions into AI for years now. With recent results from experimental AI at Expressive Intelligence Studio, they are getting closer. Emotional AI in gaming is on the horizon, promising an immersive and captivating gaming experience. It won’t be long after they succeed that we could see these AI in games.