AI is coming for your PC games, but you should be excited, not worried

Over the past week, the tech community has experienced a saturation of AI-related developments, ranging from ChatGPT to Google Bard. This surge is not unwarranted. While trends like NFTs and web3 may be transient, the presence of AI is enduring, extending even into your PC gaming experiences.However, it’s not a constant state of negativity. AI and machine learning have already demonstrated their immense utility in the realm of PC gaming, carrying significant implications for the creation and enjoyment of games. I’m not forcing unrelated ideas together; by staying with me, you’ll soon grasp the reasons behind this.

Nvidia’s Deep Learning Super Sampling (DLSS) stands as the cornerstone of AI integration in PC gaming at present. By training an AI model to reconstruct in-game frames and subsequently rendering the game at a lower resolution, you can benefit from enhanced performance without compromising visual quality.

DLSS has reached an impressive level of proficiency. Despite competing algorithms like AMD’s FidelityFX Super Resolution (FSR), Nvidia has demonstrated the distinct advantage of AI. This fact is underscored by DLSS 3, which generates supplementary frames through AI alone. AMD likely shares this viewpoint, as evidenced by the inclusion of an untapped AI accelerator in its latest RX 7900 XTX and RX 7900 XT.

Beyond mere upscaling, AI has also made its presence felt in Nvidia’s RTX Remix modding platform. This platform offers an array of AI tools designed to enhance textures, introduce additional lighting to scenes, and more. Portal RTX serves as a compelling illustration of how these AI tools can breathe new life into an aging game.

Nvidia isn’t the sole player in this arena either. A recent modification for the classic game Morrowind introduced AI-generated voice lines. Given the extensive amount of dialogue in Morrowind, achieving this would require either years of effort from modders to hire voice actors (a doubtful prospect) or an exorbitant amount of funding from Bethesda (an even more improbable scenario).

These examples only scratch the surface of how AI is currently enriching gaming experiences. However, it’s crucial to also acknowledge the fundamental role of AI within games themselves. Tools like Unreal Engine 5’s MassEntity hold the promise of effortlessly generating large quantities of AI-driven characters with minimal inputs. This exemplifies one of the most thrilling aspects of AI in PC gaming—the acceleration of game development processes.

In 2020, Nvidia accomplished the remarkable feat of training an AI model to replicate Pac-Man. Through exposure to 50,000 gameplay episodes, two neural networks collaborated to recreate a playable rendition of Pac-Man without relying on an underlying game engine. Devoid of predefined rules or parameters, the AI simply absorbed extensive gameplay and deduced how to replicate it.

This accomplishment serves as a testament to the significant strides AI has made in expediting game development. Yet, there’s no need for concern that AI will entirely take over PC game creation. As Take-Two CEO Strauss Zelnick highlighted in a recent earnings call, “[AI is] not going to allow someone to say, ‘Please develop a competitor to Grand Theft Auto that surpasses Grand Theft Auto’… people might attempt it, but it’s unlikely to succeed.”

While such an outcome might remain improbable, Zelnick was explicit about AI’s substantial role in game development. One burgeoning application in recent years is AI-driven Quality Assurance (QA) testing. Players typically invest thousands of hours in revealing only a fraction of bugs in PC games, yet AI can execute thousands of tests on a game within seconds.

Furthermore, AI holds the potential to craft more intricate and diverse games, especially those with branching narratives. AI Dungeon 2, introduced in 2019, exemplifies this by generating text-based adventures with limitless storylines. Additionally, there’s the continuous discourse around procedural generation seen in games like No Man’s Sky, where algorithms spawn billions of planets.

While games such as No Man’s Sky forge novel worlds, their origins are rooted in human design. Generative AI, on the other hand, opens doors to entirely fresh encounters devoid of rigid constraints, necessitating only adequate training.

A recent example is my interaction with the Shadows of Doubt demo—a neo-noir thriller that procedurally constructs an entire world. This experience was made possible through the Steam Next Fest. The game offers a wholly distinctive journey each time it’s played, featuring varying clues, NPCs, and settings. The infusion of AI into this context is undeniably exhilarating, as it delivers unparalleled experiences to each player engaging with the game


This Post Has One Comment

Leave a Reply