Starship Simulator Dev Dan Govier on Fleetyard AI Tools

Starship Simulator, the co-op multiplayer project from Fleetyard Studios that tasks players with crewing a massive spaceship to explore a Milky Way–inspired, real-world rendering, still doesn’t have a release date—even after a successful Kickstarter campaign in 2024 and ongoing progress on its development build. This week, lead developer Dan Govier also shared a detailed stance on how the studio is approaching AI development tools.

Govier outlines Fleetyard’s AI tool policy

In a new message explaining the studio’s thinking, Govier framed AI as a productivity aid rather than a direct measure of whether “AI was used” during game creation. He argued the more important question is how much of the final work was produced with AI versus created by humans, adding that AI tools can help smaller studios keep up with larger competitors.

Govier says Fleetyard is considering AI tools primarily to streamline workflow and remain competitive with the wider industry.
He argues that for indie teams, AI can help bridge the gap to bigger studios, while financially pressured larger studios may rely on AI tools to continue making games.
He states the studio’s current approach includes using AI for certain production steps while reserving final content for human-crafted assets.

What Fleetyard plans to use AI for (and what it won’t)

2-D, 3-D, and audio assets generated with “gen AI” are intended only as placeholders until human-created replacements are ready.
The studio is considering using an in-game LLM for NPC crewmate interactions, but says it will require several prerequisites regarding what the model is trained on.
Fleetyard says it is already using “coding agents” for “boring, time-consuming” tasks, and Govier calls them a necessity for studios that want to compete going forward.

Govier’s warning is direct: he claims any studio not leveraging coding agents will fall behind those that do. He also argues that simply writing code syntax is no longer the differentiator, and that the advantage is in designing the best solutions.

Model choice and environmental claims

Govier says Fleetyard uses models developed by Anthropic exclusively, citing a lower environmental impact. He referenced Anthropic’s reported “hydro-efficient” approach tied to a partnership with Amazon Web Services and also pointed to commitments from AWS and Microsoft to be water-positive by 2030.

The post also notes that such promises can conflict with reporting that highlights rising AI datacenter impacts on resources like water and power, based on data from Consumer Reports.

Development continues; third tech demo is out

While no release date has been announced, Fleetyard’s development build continues to move forward via its monthly updates. The studio has also released a third technology demo, following a second tech demo that was available in November 2023.