An Indian avatar is in the world’s first AI beauty pageant shortlist, where two of the four judges are also AI-powered. Mint explores the pros and cons of AI’s growing influence, from AI CEOs and assistants to realistic AI deepfakes and voice clones.
Who is the Miss AI candidate from India?
Zara Shatavari, who has over 7,100 followers on Instagram, promotes health and fashion. She is an AI-generated avatar like the other nine finalists at the World AI Creator Awards (WAICA) Miss AI Competition. Jasmeet Singh, director and co-founder of Digimozo eServices that “employs” Zara, posted on LinkedIn that AI influencers can help small brands and advertising agencies promote their products “without needing Hollywood-like budgets”. The winner, expected to be declared by end-June, will get $5000. The top three will have access to AI mentorship programmes and PR services.
Where are AI avatars being used?
They work as virtual assistants, tutors, and in video games. They serve as social media influencers, offer personalized marketing interactions, assist in patient monitoring and therapy, simulate corporate training scenarios, and serve as companions to “lonely” people. AI avatars also facilitate personalized shopping, participate in virtual meetings, and even serve as CEOs. For instance, NetDragon Websoft, Dictador Holdings, and Abu Dhabi’s International Holding Co. have all appointed AI-powered entities to executive roles, with Deep Knowledge Ventures using AI as a CEO aide.
How big is the AI avatar market?
The AI avatar (or digital human) market, valued at $4.83 billion in 2022, may touch $67.54 billion by 2032, according to Market Research Future, with the entertainment segment accounting for 23.2%. Gartner predicts avatars leveraging text-to-video via generative AI tech will support 70% of digital and marketing communications by 2025.
How can I create an AI avatar?
There are many 2D, 3D and humanoid free and paid avatar generators including Meta, Microsoft, Nvidia, Epic Games, etc. Miss AI competition partner Fanvue, too, allows influencers to create their AI versions. Firms use techniques such as generative adversarial networks, variational autoencoders, and natural language processing to help the avatar understand and respond to human language. One can integrate speech recognition capabilities using application programming interfaces.
Where’s all of this leading us?
AI’s lack of emotional intelligence and complex decision-making skills may not make it a good CEO but it can certainly assist CEOs with data-driven insights. On the flip side, deepfake videos, images and audio are making it extremely hard for humans to tell real from fake or synthetic. Moreover, if AI avatars look and sound too much like their human counterparts, a phenomenon called the ‘Uncanny Valley’, it can evoke unease or discomfort, and backfire with feelings of human mistrust and frustration.