Kevin Alster, Learning Strategist at Synthesia, talks about AI-generated avatars in learning, including technology behind them, impact on elearning, success stories, risks, and challenges. Discussion covers inclusivity, evolution of L&D with AI, and addressing ethical concerns. Episode also introduces AI avatars, replacing human hosts.
AI avatars streamline video creation, enhancing accessibility and user experience.
Technology behind AI avatars evolves to capture realistic facial expressions beyond mouth movements.
Inclusivity in avatar representation is achieved through diverse human models, ensuring relatable learning experiences.
Deep dives
AI-Generated Avatars in Learning Courses
AI-generated avatars, like those used by Synthhesia, are transforming the learning experience by simplifying video creation. By eliminating the need for actors, cameras, and studio setups, avatars streamline content production, making it as easy as creating a PowerPoint. This innovation allows for personalized video creation directly from a web browser, enhancing accessibility and user experience.
Advancements in AI Technology for Avatar Creation
The AI technology behind generating avatars involves capturing live footage of human models in a studio, reading a script. This footage is processed through AI algorithms to create models that interpret and synchronize facial movements with spoken text. The evolving technology now focuses on capturing facial and nonverbal expressions beyond just mouth movements, adding realism and immersion to avatar representations.
Personalization and Diversity in Avatar Representation
To ensure inclusivity and cultural diversity in avatar representation, Synthhesia's process involves human models recording footage to create avatars, reflecting a range of ages, backgrounds, and cultural identities. By actively seeking feedback from customers and incorporating diverse models, the avatars aim to represent real-world diversity, offering a broad range of avatars for personalized and relatable learning experiences.
Using AI Avatars for Ongoing Learning Campaigns
Managers are trained through in-person sessions and AI-based ongoing learning campaigns. The campaign involves receiving weekly customer scenarios via AI avatars to practice providing feedback. This approach encourages experiential learning and user choice, emphasizing the importance of immersion and failure in learning experiences.
Challenges and Future of AI Avatar Technology
Current limitations in AI avatar technology include restricted camera framing and educational barriers for users in video creation. Future advancements aim to introduce volumetric capture for diverse camera angles. Educational initiatives like the Deja Academy focus on enhancing video literacy and training effectiveness. Addressing these challenges involves a blend of technological innovation and educational support to optimize AI avatar utilization.
This episode is part 2 of a podcast swap with the Video Learning Lab podcast run by Kevin Alster. Kevin is a Learning Strategist at Synthesia, the Denmark-based video learning company, and was previously employed as a designer of learning experiences at The New York Times and Sotheby’s. John talks to him about Synthesia's AI-generated avatars, which are rapidly gaining use in learning programmes. This wide-ranging discussion covers many of the most topical questions about use of AI in learning. And in a first for the Learning Hack podcast, the episode is introduced by AI generated avatars, replacing the usual cast of human presenters.
0:00 - Intro
3:08 - Technology behind Synthesia's AI
11:15 - Ensuring inclusive avatar representation
20:53 - The impact on elearning
27:54 - Success stories in AI-assisted learning
35:47 - How L&D people will evolve with AI-generated Avatars