Synthesia created an interactive AI avatar of a TechCrunch journalist
Startup Synthesia has created an interactive AI avatar of a TechCrunch journalist, trained on her specific article and able to answer questions about it. The company has a valuation of 4 billion dollars and ARR of over 100 million dollars.
Synthesia, a startup specializing in creating AI video avatars, has created an interactive digital avatar of a TechCrunch journalist. According to the article, this is the first such avatar the company has created for a journalist outside its own employee, Alexandra Voicu. The avatar was trained on a specific article by the journalist about fraud at venture-capital-funded startups and only answers questions related to the content of that text — for personal questions, it redirects the questioner back to the article's topic.
Technically, the avatar combines several models: a speech-to-text model, an agentic language model processing meaning and controlling responses, a text-to-speech model, and a video model animating the speaking figure. According to the article, Synthesia uses its own video and voice models, but also allows customers to choose alternatives from Cartesia, ElevenLabs, Google, or OpenAI. Creating the avatars (a personal version for reading a script as well as an interactive version) took the Synthesia team several days and involved a photo shoot and voice recording.
Synthesia, originally a British startup, reportedly reached a valuation of 4 billion dollars this year and exceeded annual recurring revenue (ARR) of over 100 million dollars last year. Besides its classic platform for creating videos with avatars, the company also offers an agentic platform called Sessions for roleplay and surveys (including a new product, Roleplay Sessions, for practicing sales pitches) and an API platform for combining its video and voice models with other services.
The article's author describes mixed reactions from people around her to the avatar — some found it impressive, others found it unsettling ("creepy"). She also reflects on whether AI avatars could in the future supplement or replace journalists, stating that in her view, trust — as a key component of journalism — should not be outsourced to AI.
Why it matters
The case shows how far interactive AI avatar technology combining speech recognition, language models, and voice and video synthesis has advanced — and that it is available outside Synthesia's own internal use, specifically for PR purposes in the media. At the same time, it raises the question of how suitable such solutions are in areas where trust in human communication is key, such as journalism.
Two audiences, two different impacts
What this means
For individuals
It turns out that an AI clone of a specific person can faithfully mimic both voice and appearance and answer questions about a selected text, which raises questions about trust in journalism and the possibility of replacing human communication with automation.
For a business
Companies can use similar interactive avatars for PR communication, customer service, or internal training (e.g. Synthesia's Roleplay Sessions product for practicing sales conversations), which could reduce the costs of repeated human interaction.
Sales and marketingCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.