Nearly half of people in a blind video test could not tell Tavus Griffin from a real human. The result lands in the middle of remote work, hiring and sales, where face-to-face trust is assumed. Griffin is presented as the first face-to-face human interaction model. The 48 percent confusion rate is not a lab curiosity. It is a trust inflection point.
💡 Frequently Asked Questions (FAQ)
- Q: What is Tavus Griffin?
- A: Tavus Griffin is presented as the first face-to-face human interaction model, an AI designed to simulate real human video conversations.
- Q: What percentage of people could not tell Griffin from a real human?
- A: Nearly 48% of people in a blind video test could not tell Tavus Griffin from a real human.
- Q: Why does the 48% confusion rate matter?
- A: It matters because it impacts trust in remote work, hiring and sales where face-to-face authenticity is assumed, creating a trust inflection point.
- Q: Is the Griffin test result just a lab curiosity?
- A: No, the result is described as a trust inflection point with real implications for how people verify identity in video interactions.