The AI coworkers in this game are voiced by a machine (TTS), and their dialogue was really written by AI together with self portraits, on purpose and under human review. The slop is diegetic. The code was written with AI agents and reviewed by the last human in that loop, the developer. That is also the plot. No AI runs at runtime. Everything was baked and reviewed before shipping.
Art direction steered asset creation, SFX, store capsule artwork and the logo were produced with generative AI assistance then hand-edited. No generative AI runs at runtime; the game generates no content live.
Generative AI is used to assist with the translation and localization of in-game text and Steam store page content from the original Japanese into other supported languages.
Generative AI was used during development to assist in creating some pre-generated visual artwork and illustrative assets, including artwork used within the Chronicle system. All AI-assisted content is created and reviewed during development and is included in the game as finished assets. Echo Harvest does not use generative AI to create content dynamically during gameplay.
Environments and environmental artwork are not created using generative AI.
Generative AI is used as part of the early development workflow for prototyping some creature artwork, which is going to be replaced during development.
We used AI tools in two places during development, and in neither of them to replace creative work.
Our French, Spanish, Italian, Japanese and Korean localisations are made using AI. We also used AI assistance while writing and restructuring parts of our codebase.
Everything players see and hear — art, animation, sound, music, game design and all original writing — was made by people.
Generative AI is used in a supplementary capacity, at two points. First, as a drafting step for some of the game's 2D illustrations: every such asset is then rebuilt through our own production pipeline — reduced to a strict 1-bit black-and-white format, curated, and animated frame by frame with in-house tools — so that it matches our art direction. Second, the game is written in French, and its translations into the eight other supported languages are AI-assisted. No generative AI runs in the game itself: all content ships fixed with the build.
Some of the 2D artwork in this game — character portraits, monster illustrations, board tiles, backgrounds and store art — was created with the help of generative AI image tools, then reviewed, edited and assembled by the developer. All generated assets were checked before shipping. The game does not generate any content live during play: everything is pre-rendered and included in the build. Game design, rules, code, text and level layout were authored by the developer.
This game was created by an independent developer. Generative AI tools were used during development to assist with programming, debugging, text editing, and the initial localization of Simplified Chinese, Traditional Chinese, English, Japanese, and Korean. All AI-assisted content was manually reviewed, edited, and approved by the developer. The game does not use generative AI to create content during gameplay, and player data is not used for AI training.
Swirl Together contains a rotoscoping/masking model similar to Adobe’s roto brush, or Davinci Resolve's magic masking tool, and a depth-estimating model for colliders.
Generative AI technologies were utilized during the development and content creation process of our game. In addition to software development support, AI tools were used to draft in-game narratives, assist with localization, and generate store page descriptions. All AI-assisted content has been thoroughly reviewed, edited, and curated by the development team to ensure quality standards and compliance with copyright policies.
Space Travelers uses generative AI only as a writing aid during development.
The game's text content (interface text, event and log lines) and its
localization into the 14 supported interface languages were produced with
the assistance of large language models, directed and reviewed by the
developer. Parts of this store page were drafted the same way.
No AI-generated images, audio or video are used: all artwork, animation
and sound in the game are produced by hand-written procedural code. The
game does not generate any content with AI while you play, and it never
connects to an external AI service.
Artificial Intelligence Usage Notice:
We have used some generative models to prototype new ideas, and some of them have been made into production. However, AI-generated content is no more than 1% of the total game assets and arts.
Some sprites started as AI-generated references before being rebuilt in the game's palette and resolution. All English writing is my own; the other nine languages are machine translated. No AI runs in the game itself. Nothing is generated while you play, and it never connects to an AI service.
Some of Card Slay's pre-generated content was created with the help of generative AI tools and then reviewed and finished by hand. No content is generated by AI during gameplay.
Some pre-rendered assets were created with generative AI tools and then curated, edited and integrated by hand: music tracks and their cover art, some sound effects and portions of 2d artwork (weapon, prop, and enemy sprites.) All generated content was reviewed by the developer before inclusion. No content is AI-generated at runtime.
AI (WAN 2.2) was used to help generate idle animation frames from 100% original, human-made art. All animations were reviewed and curated by the developer before being included in the game.
The game is developed entirely by a solo developer. AI has therefore only been leveraged as a tool for productivity and efficiency - the overall game is fully imagined, designed, and brought together by the developer.
Generative AI was used to create selected store and library artwork, music, and sound-effect source material. The outputs were curated and edited by the developer before inclusion. No generative AI runs during gameplay, and players cannot generate AI content.
Standard C++ game code, written during development with the assistance of an AI programming assistant, used as a pair-programming tool. This includes gameplay systems (crafting, inventory, world simulation), rendering and UI code, and development tooling. It is ordinary, static game logic.
AI assistance was used only during development, directed by myself through specific requests, with every change reviewed, tested, and committed to version control by myself.
The shipped game contains no AI models and does not execute or generate any code or content via AI at runtime. The game is a fully offline, single-player experience.
AI was used to assist with writing the game source code.
All visual assets are handmade pixel art by the author.
All audio assets are licensed from 3rd party sample banks.
Pixel art assets (characters, monsters, bosses, projectiles, spirits, tiles) were created with PixelLab, an AI pixel-art tool, and then edited by hand to fit the game. No AI is used at runtime. All code, game design, sound effects and music are human-authored (audio is synthesized in code).
This game uses generative AI in two ways. (1) Live generation: narrative text and NPC dialogue are generated in real time based on the player's choices and current game state. Illustrative images for locations, scenes and items are also generated on demand the first time such an element appears, then cached and reused. Core game rules, numerical outcomes, inventory changes, and progression are always resolved by deterministic game systems. Live text generation is governed by server-side content rules that cannot be directly modified by player input. Live image generation uses server-controlled prompt templates based on structured game data (entity type, name, and a short in-world description) within a fixed ink-painting style; players cannot directly provide image-generation prompts. (2) Development-time generation: some static art assets (including location, scene, item and background illustrations) were generated with image-generation AI tools during development. Store-facing marketing assets and selected screenshots containing such artwork were reviewed by a human before publication. Store page copy was drafted with AI assistance and reviewed by a human.
本游戏以两种方式使用生成式AI。(1)实时生成:叙事文本与NPC对话根据玩家选择与当前游戏状态实时生成;地点/场景/物品的插画也会在其首次出现时按需生成,之后缓存复用。核心规则、数值结算、物品变动与进度始终由确定性游戏系统裁定。实时文本生成受服务端内容规则约束(玩家输入无法直接修改);实时图像生成使用服务端控制的提示词模板,基于结构化游戏数据(实体类别、名称与简短的世界内描述)并锁定水墨画风格,玩家不能直接书写图像提示词。(2)开发期生成:部分静态美术素材(包括地点、场景、物品与背景插画)在开发过程中使用图像生成AI制作;用于商店宣传以及所选商店截图中的相关素材均经过人工审阅后发布。商店页文案由AI辅助创作、经人工审定。
Some promotional and store assets for this game, including the capsule images, library artwork and application icons, were created with the assistance of generative AI tools.
All in game visuals are generated procedurally by the game's own code at runtime: characters, buildings, terrain, animals and effects are drawn from algorithmic vector shapes, not from AI generated images. All audio is synthesised in real time by the game engine. All game code was written by the developer.
Ashen Wild was created almost exclusively with AI tools to generate the code and graphics, VFX, sound etc. The store page, website and other marketing materials were created using AI.
Some of this game's art and audio assets were created using generative AI tools, from the studio's own prompts and under our tool licenses, then curated and edited by the team. All game code and text were written by hand.
Parts of this game's content were created with generative AI tools:
- 3D models and their textures (towers, enemies, terrain and scenery props) were generated with Meshy AI, then hand-adjusted and assembled by the developer.
- The English localization was AI-translated from the original Chinese text and then reviewed by the developer.
- Some in-game flavor text (event, relic and achievement descriptions) was drafted with an AI writing assistant.
- Some store page and marketing artwork was produced with AI image tools.
All of this content is generated ahead of time and ships as fixed assets. The game does not generate any content live during play and does not connect to any AI service at runtime. Sound effects and music are procedurally synthesized by our own scripts and are not AI-generated.
Some art assets, localized text, and store marketing materials were created with the assistance of generative AI during development, then reviewed, edited, and integrated by the developer. The game does not use generative AI to create content during gameplay.
As a solo developer, AI has been a big help in making the development of this game more efficient by assisting in, but not be the sole performer of: testing, performing repetitive and some programming tasks, brainstorming and discussions, translations and some design.
This game uses generative AI tools to help create static visual assets, including card illustrations and background artwork. All AI-generated content is produced during development and does not change during gameplay — nothing is generated in real time.
I used generative AI in the process of creating most art/visual assets in Trouble Shooters, with touch ups made after generation. Music and sound effects were either composed without the use of generative AI or sourced from sound libraries that predate generative AI workflows. Nothing is generated on-the-fly.
Some in-game text (such as the messages in bottles) and store page copy were written with the assistance of AI tools. All content has been reviewed and edited by the developer.
Generative AI was used as an auxiliary tool during development and to make minor edits and compositions to graphical assets, including logos. No generative AI is used to generate content dynamically during gameplay.
3D models for props and buildings have been generated with code developed by AI. The game splash screens and logo have been generated by AI. Music is generated live with code developed by AI.
During development, we used AI tools to assist in creating some art assets, voice acting, and dialogue. All AI-assisted content is pre-generated before release — the game does not generate content with AI in real time while you play. All such content is reviewed and refined by our team to ensure quality and originality, and to avoid generic-looking characters.
The AI coworkers in this game are voiced by a machine (TTS), and their dialogue was really written by AI together with self portraits, on purpose and under human review. The slop is diegetic. The code was written with AI agents and reviewed by the last human in that loop, the developer. That is also the plot. No AI runs at runtime. Everything was baked and reviewed before shipping.
Art direction steered asset creation, SFX, store capsule artwork and the logo were produced with generative AI assistance then hand-edited. No generative AI runs at runtime; the game generates no content live.
Generative AI is used to assist with the translation and localization of in-game text and Steam store page content from the original Japanese into other supported languages.
Generative AI was used during development to assist in creating some pre-generated visual artwork and illustrative assets, including artwork used within the Chronicle system. All AI-assisted content is created and reviewed during development and is included in the game as finished assets. Echo Harvest does not use generative AI to create content dynamically during gameplay.
Environments and environmental artwork are not created using generative AI.
Generative AI is used as part of the early development workflow for prototyping some creature artwork, which is going to be replaced during development.
We used AI tools in two places during development, and in neither of them to replace creative work.
Our French, Spanish, Italian, Japanese and Korean localisations are made using AI. We also used AI assistance while writing and restructuring parts of our codebase.
Everything players see and hear — art, animation, sound, music, game design and all original writing — was made by people.
Generative AI is used in a supplementary capacity, at two points. First, as a drafting step for some of the game's 2D illustrations: every such asset is then rebuilt through our own production pipeline — reduced to a strict 1-bit black-and-white format, curated, and animated frame by frame with in-house tools — so that it matches our art direction. Second, the game is written in French, and its translations into the eight other supported languages are AI-assisted. No generative AI runs in the game itself: all content ships fixed with the build.
Some of the 2D artwork in this game — character portraits, monster illustrations, board tiles, backgrounds and store art — was created with the help of generative AI image tools, then reviewed, edited and assembled by the developer. All generated assets were checked before shipping. The game does not generate any content live during play: everything is pre-rendered and included in the build. Game design, rules, code, text and level layout were authored by the developer.
This game was created by an independent developer. Generative AI tools were used during development to assist with programming, debugging, text editing, and the initial localization of Simplified Chinese, Traditional Chinese, English, Japanese, and Korean. All AI-assisted content was manually reviewed, edited, and approved by the developer. The game does not use generative AI to create content during gameplay, and player data is not used for AI training.
Swirl Together contains a rotoscoping/masking model similar to Adobe’s roto brush, or Davinci Resolve's magic masking tool, and a depth-estimating model for colliders.
Generative AI technologies were utilized during the development and content creation process of our game. In addition to software development support, AI tools were used to draft in-game narratives, assist with localization, and generate store page descriptions. All AI-assisted content has been thoroughly reviewed, edited, and curated by the development team to ensure quality standards and compliance with copyright policies.
Space Travelers uses generative AI only as a writing aid during development.
The game's text content (interface text, event and log lines) and its
localization into the 14 supported interface languages were produced with
the assistance of large language models, directed and reviewed by the
developer. Parts of this store page were drafted the same way.
No AI-generated images, audio or video are used: all artwork, animation
and sound in the game are produced by hand-written procedural code. The
game does not generate any content with AI while you play, and it never
connects to an external AI service.
Artificial Intelligence Usage Notice:
We have used some generative models to prototype new ideas, and some of them have been made into production. However, AI-generated content is no more than 1% of the total game assets and arts.
Some sprites started as AI-generated references before being rebuilt in the game's palette and resolution. All English writing is my own; the other nine languages are machine translated. No AI runs in the game itself. Nothing is generated while you play, and it never connects to an AI service.
Some of Card Slay's pre-generated content was created with the help of generative AI tools and then reviewed and finished by hand. No content is generated by AI during gameplay.
Some pre-rendered assets were created with generative AI tools and then curated, edited and integrated by hand: music tracks and their cover art, some sound effects and portions of 2d artwork (weapon, prop, and enemy sprites.) All generated content was reviewed by the developer before inclusion. No content is AI-generated at runtime.
AI (WAN 2.2) was used to help generate idle animation frames from 100% original, human-made art. All animations were reviewed and curated by the developer before being included in the game.
The game is developed entirely by a solo developer. AI has therefore only been leveraged as a tool for productivity and efficiency - the overall game is fully imagined, designed, and brought together by the developer.
Generative AI was used to create selected store and library artwork, music, and sound-effect source material. The outputs were curated and edited by the developer before inclusion. No generative AI runs during gameplay, and players cannot generate AI content.
Standard C++ game code, written during development with the assistance of an AI programming assistant, used as a pair-programming tool. This includes gameplay systems (crafting, inventory, world simulation), rendering and UI code, and development tooling. It is ordinary, static game logic.
AI assistance was used only during development, directed by myself through specific requests, with every change reviewed, tested, and committed to version control by myself.
The shipped game contains no AI models and does not execute or generate any code or content via AI at runtime. The game is a fully offline, single-player experience.
AI was used to assist with writing the game source code.
All visual assets are handmade pixel art by the author.
All audio assets are licensed from 3rd party sample banks.
Pixel art assets (characters, monsters, bosses, projectiles, spirits, tiles) were created with PixelLab, an AI pixel-art tool, and then edited by hand to fit the game. No AI is used at runtime. All code, game design, sound effects and music are human-authored (audio is synthesized in code).
This game uses generative AI in two ways. (1) Live generation: narrative text and NPC dialogue are generated in real time based on the player's choices and current game state. Illustrative images for locations, scenes and items are also generated on demand the first time such an element appears, then cached and reused. Core game rules, numerical outcomes, inventory changes, and progression are always resolved by deterministic game systems. Live text generation is governed by server-side content rules that cannot be directly modified by player input. Live image generation uses server-controlled prompt templates based on structured game data (entity type, name, and a short in-world description) within a fixed ink-painting style; players cannot directly provide image-generation prompts. (2) Development-time generation: some static art assets (including location, scene, item and background illustrations) were generated with image-generation AI tools during development. Store-facing marketing assets and selected screenshots containing such artwork were reviewed by a human before publication. Store page copy was drafted with AI assistance and reviewed by a human.
本游戏以两种方式使用生成式AI。(1)实时生成:叙事文本与NPC对话根据玩家选择与当前游戏状态实时生成;地点/场景/物品的插画也会在其首次出现时按需生成,之后缓存复用。核心规则、数值结算、物品变动与进度始终由确定性游戏系统裁定。实时文本生成受服务端内容规则约束(玩家输入无法直接修改);实时图像生成使用服务端控制的提示词模板,基于结构化游戏数据(实体类别、名称与简短的世界内描述)并锁定水墨画风格,玩家不能直接书写图像提示词。(2)开发期生成:部分静态美术素材(包括地点、场景、物品与背景插画)在开发过程中使用图像生成AI制作;用于商店宣传以及所选商店截图中的相关素材均经过人工审阅后发布。商店页文案由AI辅助创作、经人工审定。
Some promotional and store assets for this game, including the capsule images, library artwork and application icons, were created with the assistance of generative AI tools.
All in game visuals are generated procedurally by the game's own code at runtime: characters, buildings, terrain, animals and effects are drawn from algorithmic vector shapes, not from AI generated images. All audio is synthesised in real time by the game engine. All game code was written by the developer.
Ashen Wild was created almost exclusively with AI tools to generate the code and graphics, VFX, sound etc. The store page, website and other marketing materials were created using AI.
Some of this game's art and audio assets were created using generative AI tools, from the studio's own prompts and under our tool licenses, then curated and edited by the team. All game code and text were written by hand.
Parts of this game's content were created with generative AI tools:
- 3D models and their textures (towers, enemies, terrain and scenery props) were generated with Meshy AI, then hand-adjusted and assembled by the developer.
- The English localization was AI-translated from the original Chinese text and then reviewed by the developer.
- Some in-game flavor text (event, relic and achievement descriptions) was drafted with an AI writing assistant.
- Some store page and marketing artwork was produced with AI image tools.
All of this content is generated ahead of time and ships as fixed assets. The game does not generate any content live during play and does not connect to any AI service at runtime. Sound effects and music are procedurally synthesized by our own scripts and are not AI-generated.
Some art assets, localized text, and store marketing materials were created with the assistance of generative AI during development, then reviewed, edited, and integrated by the developer. The game does not use generative AI to create content during gameplay.
As a solo developer, AI has been a big help in making the development of this game more efficient by assisting in, but not be the sole performer of: testing, performing repetitive and some programming tasks, brainstorming and discussions, translations and some design.
This game uses generative AI tools to help create static visual assets, including card illustrations and background artwork. All AI-generated content is produced during development and does not change during gameplay — nothing is generated in real time.
I used generative AI in the process of creating most art/visual assets in Trouble Shooters, with touch ups made after generation. Music and sound effects were either composed without the use of generative AI or sourced from sound libraries that predate generative AI workflows. Nothing is generated on-the-fly.
Some in-game text (such as the messages in bottles) and store page copy were written with the assistance of AI tools. All content has been reviewed and edited by the developer.
Generative AI was used as an auxiliary tool during development and to make minor edits and compositions to graphical assets, including logos. No generative AI is used to generate content dynamically during gameplay.
3D models for props and buildings have been generated with code developed by AI. The game splash screens and logo have been generated by AI. Music is generated live with code developed by AI.
During development, we used AI tools to assist in creating some art assets, voice acting, and dialogue. All AI-assisted content is pre-generated before release — the game does not generate content with AI in real time while you play. All such content is reviewed and refined by our team to ensure quality and originality, and to avoid generic-looking characters.