Across the Jalwa Lottery industry, new generative AI models are constantly claiming to fundamentally alter how video games are created. Despite such claims, however, many of these new technologies have significant limitations that prevent them from making any substantial changes to existing gameplay.

One of the key challenges with gaming AI is that it must be able to learn and adapt to its player’s skill level. AI that can flexibly adjust difficulty levels, pacing, puzzles and combat encounters to keep the game fresh would be a massive improvement over the rote and predictable experiences of today.

Exploring the Limits of AI Autonomy in Games

Other challenges include the need for AI to behave more naturally, something that has largely been limited by the need for game designers to pre-program its actions. Recently, however, some generative AI programs have been able to create more realistic behaviors by learning from examples. This is most apparent in the augmented reality drawing app Dots and Boxes, where players are asked to draw a specific object. The program then compares the player’s drawings to previous examples in order to guess correctly what the player is trying to draw.

In addition, generative AI can be used to create smarter NPCs in video games that are less predictable and more natural. A student at Tokyo’s Rikkyo University, for example, has developed an AI that takes speech as input and generates body gesture output to create more realistic characters. It’s a development that demonstrates how the combination of generative AI and deep neural networks can create characters that feel more real.