Pillar 01 · Research Program
Voice-driven dialogue agents, AI-generated social media, and cinematic animations that make opera discoverable and emotionally compelling for new audiences.
A central challenge in opera audience development is not access — recordings, streaming, and digital archives are widely available — but discoverability and emotional entry. People who have never experienced opera lack the vocabulary, context, and community to begin. This research initiative addresses that gap directly, using AI as a bridge.
We develop and evaluate voice-driven dialogue systems that respond naturally to questions about operatic works, singers, history, and craft. Our research examines how natural-sounding voice synthesis combined with large language models shapes learner engagement, comprehension, and willingness to attend live performances.
A parallel content development stream uses generative AI to produce social media posts, short documentary segments, and cinematic animations that convey emotionally resonant messages about opera — reaching platforms and audiences that opera institutions have historically been unable to access.
Our guiding hypothesis: AI-generated content and conversational agents can measurably shift awareness, sentiment, and attendance intent among non-traditional opera audiences, particularly adults aged 18–45. We test this through controlled studies of engagement metrics, pre/post surveys of audience intent, and longitudinal attendance tracking at partner institutions through the STAGE data platform.
Voice synthesis technology — including high-fidelity AI voice from ElevenLabs — is central to both the conversational agent and content production research tracks. Voice delivery creates an experience that feels personal and immediate, rather than transactional or text-based.
Results from this research will be published in peer-reviewed venues and contribute open frameworks and datasets for the broader research community.
Developing and benchmarking voice-driven dialogue agents for opera education. Research questions include optimal conversation design for cultural domains, factual grounding, handling of user misconceptions, and the role of voice persona in building rapport and trust with first-time listeners.
Designing and evaluating speech-based interfaces that allow students and the public to navigate complex cultural knowledge through natural conversation. Studies compare learning outcomes and engagement depth against text-based equivalents, examining the unique affordances of voice.
Producing and A/B testing AI-generated posts, short-form video scripts, and visual content about opera across social platforms. Research identifies content formats, narrative frames, and emotional approaches most effective at driving awareness and attendance intent among 18–45 audiences.
Creating short cinematic films and animations using generative AI video tools to tell compelling stories about opera — its artists, history, and emotional power. Projects explore how AI-produced narrative media shifts cultural perception and attracts younger first-time attendees to live performance.