The Divinity Developer Explains Its Application of Generative AI for New Divinity

The team behind popular titles like Baldur's Gate 3 and Divinity: Original Sin recently teased its new project, sparking immense excitement within the player base. However, recent remarks from the company's co-founder have introduced a new dimension to the discussion, focusing on the studio's philosophy toward AI tools.

Augmenting Workflows, Not Cutting Jobs

In a recent message, Larian's director detailed that the company is utilizing machine learning for specific supporting functions. These involve enhancing pitch decks, generating initial artistic references, and writing temporary text.

Importantly, Vincke made clear that the final assets in the game will be crafted entirely by human creatives. "Larian is creating everything in-house," he said.

Larian is constantly expanding our team of writers and are busily assembling writing teams.

Since visual development is being explicitly mentioned — we currently have twenty-three visual developers and have positions available for additional creatives.

Each initiative we do is additive and aimed at having people spend greater focus on actual creation.

Any AI system implemented properly is additive to a developer's process, not a replacement for their craft.

Tempering Reactions with Clear Intent

The news of AI usage at first provoked backlash among some the community. In response, Vincke provided further elaboration on public forums.

"At Larian, we employ machine learning to explore references, just like we use the internet and reference books," he stated. "During the initial ideation stages we use it as a basic framework for composition which we then substitute with original illustrations."

He continued, "Larian brings on creatives for their inherent skill, not for their willingness to replicate what a algorithm proposes."

Key Areas of AI Integration

Vincke had in the past outlined the studio's focused approach to machine learning, categorizing its use into key areas:

• Streamlining Repetitive Work: Areas like refining animations, dialogue cleanup, and pipeline-specific tasks like adapting animations for different models.
• Fast-Tracked Experimentation: Using systems to rapidly prototype rough versions of scenarios to validate concepts ahead of complete development.
• Future Potential for Gameplay: Exploring how machine learning could one day facilitate new forms of gameplay, particularly in simulating dynamic reactions in a complex RPG.

He specifically noted that central narrative domains — like visual art — are are absolutely not areas where the studio is cutting human involvement. On the contrary, Larian is expanding its staff in these precise roles.

"We are not releasing a game with AI-generated content, and we are certainly not planning on cutting creatives to swap them out with artificial intelligence," Vincke concluded.