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

The team behind popular RPGs like Baldur's Gate 3 and Divinity: Original Sin just teased its next major project, sparking significant anticipation within the industry. However, subsequent comments from the studio's lead designer have introduced nuance to the discussion, focusing on the developer's philosophy toward machine learning.

Augmenting Workflows, Not Cutting Jobs

In a new clarification, Swen Vincke explained that the developer is utilizing generative AI for particular ancillary purposes. These include fleshing out pitch decks, producing early-stage artistic references, and drafting draft copy.

Importantly, Vincke made clear that the end assets in the game will be crafted entirely by real writers. "Larian is developing all the content in-house," he affirmed.

Our studio is actively increasing our roster of writers and are actively putting together dedicated writer rooms.

Given that concept art is being particularly called out — we currently have 23 concept artists and have job openings for additional artists.

All our efforts we do is additive and focused on enabling creatives to spend more time on making content.

Any machine learning application applied correctly is a boost to a developer's workflow, not a replacement for their talent.

Responding to Feedback and Defining the Path

The news of AI usage at first generated backlash among a segment of the community. In reaction, Vincke issued more elaboration on public forums.

"At Larian, we employ machine learning to explore references, just like we use Google and physical media," he stated. "In the initial ideation stages we use it as a simple sketch for layout which we then substitute with authentic illustrations."

He added, "We've hired artists for their creative vision, not for their capacity to follow what a machine suggests."

Three Pillars of Practical Application

Vincke had in the past outlined the company's targeted approach to AI and ML, categorizing its use into three main pillars:

• Handling Monotonous Jobs: Areas like refining animations, voice editing, and technical processes like adapting animations for different models.
• Fast-Tracked Experimentation: Using technology to speedily create rough versions of scenarios to experiment with concepts ahead of complete development.
• Experimental Frontiers: Investigating how machine learning could in the future create emergent player agency, specifically in creating player-driven narratives in a detailed game universe.

He specifically affirmed that core creative disciplines — like visual art — are are in no way departments where the studio is replacing human involvement. On the contrary, Larian is expanding its staff in these exact fields.

"Larian is not releasing a game with machine-made assets, and we are certainly not looking at reducing staff to replace them with artificial intelligence," Vincke concluded.