The Creators of Baldur's Gate 3 Explains Its Use of Machine Learning for Next Project

The developer behind acclaimed role-playing games like Baldur's Gate 3 and Divinity: Original Sin just unveiled its upcoming project, creating immense anticipation within the player base. However, subsequent comments from the studio's co-founder have introduced a new dimension to the narrative, focusing on the team's approach toward generative artificial intelligence.

A Tool for Ideation, Not Replacement

In a latest clarification, the studio's founder detailed that the company is using machine learning for certain supporting tasks. These involve enhancing presentation materials, generating early-stage artistic references, and creating placeholder text.

Notably, Vincke made clear that the end material in the game will be created exclusively by actual writers. "We are developing all the content manually," he stated.

Our studio is actively growing our roster of concept artists and are actively putting together writing teams.

Given that concept art is being specifically referenced — we currently have 23 visual developers and have roles to fill for further creatives.

Everything we do is supplementary and aimed at having people spend more time on making content.

Any machine learning application implemented properly is supplementary to a artist's routine, not a substitute for their skill.

Responding to Feedback and Defining the Path

The news of employing this technology originally provoked concern among some the player base. In response, Vincke issued additional clarification on online platforms.

"Our team utilizes these tools to research ideas, just like we use the internet and physical media," he explained. "In the very early ideation stages we use it as a simple sketch for composition which we then replace with hand-crafted concept art."

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

Key Areas of AI Integration

Vincke had previously broken down the company's targeted strategy to AI and ML, defining its use into three main areas:

• Automation of Tedious Tasks: Areas like polishing mocap data, audio processing, and Larian-specific work like adapting animations for different models.
• Rapid Prototyping ('White Boxing'): Using systems to quickly build rough models of gameplay ideas to experiment with concepts ahead of full development.
• Long-Term Aspirations: Exploring how machine learning could one day enhance new forms of reactivity, especially in managing dynamic reactions in a vast role-playing world.

He specifically noted that core creative disciplines — like writing — are are absolutely not fields where the team is cutting artistic input. In fact, Larian is expanding its staff in these exact roles.

"Our studio is not releasing a game with AI-generated content, nor considering trimming down teams to replace them with AI," Vincke summarized.