Synthesia, a British AI startup, has introduced groundbreaking "Expressive Avatars," boasting the most authentic emotional expressions in the market, as reported by The Next Web.

Powered by AI models trained on real actors, these digital beings blur the line between virtual and real, offering users a unique video creation experience. Using automatic sentiment prediction, the avatars can infer emotions from text, delivering performances with nuanced tones, gestures, and facial expressions. 

However, some limitations exist, such as exaggerated emotions and restricted movements. 

With plans for major improvements, including envisioning them as integral parts of the digital workplace, Synthesia's avatars may soon revolutionize various industries, potentially rivaling even human performers in authenticity and realism.

READ NEXT: AI Companies Face Data Shortage After Using Up All Available Internet Resources for Model Training

Critics of AI Avatars With Expressive Emotions

Following Synthesia's aim to expand its avatars' applications beyond its current customer base, there are concerns about their potential misuse for spreading disinformation. 

As reported by VCPost, AI avatars have been exploited to produce fake news in different regions, particularly in Meta apps. Therefore, prompting legislation to implement stricter rules and improve security measures for each individual is necessary. 

For now, there are ongoing efforts to mitigate risks. But worries persist about the avatars' ability to deceive viewers and manipulate information, especially as they become increasingly indistinguishable from real humans.

It's also uncertain where Synthesis will extract human data, considering the AI industry faces data shortage after using all available internet resources for model training.

READ MORE: Microsoft Starts Exploration On AI in Gaming Through An Upcoming Xbox AI Chatbot