AI Revives a Game Lost 20 Years Ago, Its Creator Says
The case demonstrates a practical use of AI to recover and run old game data that would otherwise be lost to hardware and software obsolescence.
Reporting from 1 sources: Game Spark.
Kaiware, a digital art engineer, used generative AI to revive "Kidou Toushu Sanshin-Oh," a self-made game from his high school days that had been unplayable for over 20 years. He fed the demo binary and data to Codex GPT5.5, which adapted it to run on Windows 11. The game plays with some slowdown and glitches but is functional, and a second revived title followed.
Kaiware, a digital art engineer, posted on X that he had revived a game he made in high school more than 20 years ago. The game, "Kidou Toushu Sanshin-Oh," had been lost after his PC broke during development. A friend had kept the demo data, but it would not run on modern systems. Kaiware gave the demo binary and data to Codex GPT5.5 and asked it to make the game work on Windows 11. The AI succeeded, though Kaiware notes slowdowns, freezes, and low-quality MIDI playback. The revived version also added a windowed mode, which the original lacked. He used the same method to bring back a second game, "Kidou Dasha Sanken-Oh."
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.