The Creators of Baldur's Gate 3 Clarifies Its Application of Machine Learning for Next Divinity

The studio behind acclaimed RPGs like Baldur's Gate 3 and Divinity: Original Sin just shown its new project, creating a wave of anticipation within the gaming community. However, recent remarks from the company's lead designer have brought a new dimension to the conversation, touching on the developer's philosophy toward AI tools.

A Tool for Ideation, Not Replacement

In a new statement, the studio's founder explained that the team is employing generative AI for certain preliminary functions. These include fleshing out pitch decks, creating initial concept art, and writing placeholder copy.

Notably, Vincke stressed that the final assets in the game will be authored exclusively by real writers. "We are creating all the content manually," he stated.

Our studio is continuously expanding our roster of concept artists and are actively putting together dedicated writer rooms.

As this area is being particularly called out — we right now have twenty-three artistic staff and have roles to fill for more talent.

All our efforts we do is additive and designed to having people spend more time on the creative process.

Any ML tool applied correctly is a boost to a developer's routine, not a substitute for their skill.

Addressing Concerns and Clarifying the Vision

The admission of employing this technology originally sparked unease among a segment of the fanbase. In reply, Vincke provided additional clarification on online platforms.

"At Larian, we employ machine learning to gather inspiration, in the same way we use the internet and reference books," he explained. "During the very early brainstorming phase we use it as a simple sketch for structure which we then swap out with original concept art."

He added, "Our studio recruits talent for their creative vision, not for their willingness to follow what a algorithm proposes."

Three Pillars of Practical Application

Vincke had in the past broken down the company's practical approach to machine learning, grouping its use into three main pillars:

• Handling Monotonous Jobs: Areas like motion capture cleaning, audio processing, and pipeline-specific tasks like adjusting assets for various species.
• Accelerated Iteration: Using tools to speedily create rough versions of mechanics to validate concepts before full implementation.
• Future Potential for Gameplay: Exploring how machine learning could eventually enhance emergent reactivity, specifically in managing dynamic reactions in a vast role-playing world.

He specifically stated that central narrative domains — such as visual art — are are absolutely not areas where the company is replacing artistic involvement. In fact, Larian is recruiting more in these exact positions.

"Larian is neither releasing a game with machine-made assets, nor planning on reducing creatives to substitute them with AI," Vincke stated definitively.