MOST：AI生成音声でクラクフのユダヤ遺産を蘇らせる

MOST — ポーランド語で「橋」を意味する — は、AIを活用した位置情報ベースのナラティブ体験で、クラクフのユダヤ文化祭で初公開されます。

歴史、場所、機械学習を横断する音響システムアーティスト、アンドリュー・メルキオールによって考案・制作されたMOSTは、AI音声合成、アーカイブ資料、位置情報ベースのストーリーテリングを組み合わせ、カジミエシュの重層的な歴史にリスナーを没入させます。

聞き、案内し、記憶するAI音声

This innovative guide draws on personal testimonies and historical documents to reveal Jewish life in interwar Poland.

As visitors move through the district, they’re met with intimate, multilingual voices that respond to location, connecting individual experience with collective memory in an immediate, location-aware format.

MOST uses ElevenLabs voice technology to animate these stories. A female Polish narrator guides listeners through Kazimierz, while the voice of a fictional Jewish male — shaped from period sources — offers playful and poignant observations in English. The result is a layered, adaptive soundscape that bridges the personal and the historical.

MOST is also a pilot for a new class of narrative engines — blending voice AI, location awareness, and cultural sovereignty to deliver adaptive, site-specific experiences across cities and generations.

Backed by ElevenLabs Impact

MOST was developed through the ElevenLabs Impact initiative — supporting cultural and civic uses of AI.

As a company founded by Polish engineers, helping bring Krakow’s Jewish past into a new era of digital remembrance is deeply meaningful to us.

Voice can do more than speak. It can remember.