AI Fan Along
Sheffield-based Joi Polloi is a double BAFTA-award winning digital agency specialising in creating compelling digital experiences, bespoke platforms and innovative products.
Joi Polloi collaborated with the University of York to explore conversational AI technology that allows children to safely engage with their favourite TV show characters through voice-assisted smart speakers such as an Amazon Alexa or Google Home.
The R&D project explored AI technology to synthesise voice, and natural language processing (NLP) to create a conversational AI model.
Together with academics from the University of the York, Joi Polloi researched a wide range of immersive audio content. A content gap was identified for children aged 11-16 who are currently in secondary school education (Key Stage 3 and 4). Particularly around experiences that engage with children through conversation as opposed to an extension of storytelling.
The AI Fan Along project has enabled Joi Polloi to understand the technical opportunities and limitations of voice-enabled services for storytelling aimed at younger audiences.
Joi Polloi has approached a number of UK and global broadcasters and production companies to explore the commercial viability of the project.
Jason Devoy, director and industry partner says:
“In the age of on-demand services and binge-watching, tuning in at a particular time to watch your favourite programme and then waiting a week for the next episode has been somewhat lost.
We wanted to explore how to rekindle that buzz and sense of drama once the credits start rolling on your favourite show. Imagine how it would feel to have the opportunity to forge a bond with your favourite character – discuss the themes and plot lines of the show and build suspense for the next episode.
Our research explored the AI technologies available to create conversational engagement and storytelling to younger audiences through smart speakers. The project leveraged research from leading academics in the field of AI at the University of York, with our commercial knowledge of the broadcasting industry.
The initial findings indicated that there is further research needed to help refine the creative concept and scripting process: alongside providing us with the experience to build and develop an authentic and realistic voice synthesis model.
We were able to develop new internal skill sets and knowledge-sharing from our relationship with the University of York’s DC Labs, alongside gaining consultancy advice from leading audio specialists and support from leading broadcasters and production companies.”