Experiencing the Future: An Interactive Digital Avatar in Journalism
When Alexandru Voica, head of corporate affairs at Synthesia—a leading video-generation startup—shared a remarkable addition to their PR team this summer, it immediately caught my attention. The innovation? An interactive virtual avatar of Alexandru himself, meticulously trained to answer common press inquiries about Synthesia’s mission and technology. Having just participated in a panel where PR professionals debated the appropriateness of AI-generated pitches, this avatar represented a quantum leap—what felt like the “final boss” of AI usage in public relations.
In September, Synthesia welcomed me to their new New York office. Originally a U.K.-based company, Synthesia has quickly become a powerhouse in the digital avatar space, standing alongside peers like D-ID, HeyGen, and Colossyan. The company’s meteoric rise is underscored by its recent achievements: reaching a $4 billion valuation earlier this year and surpassing $100 million in annual recurring revenue (ARR) last year.
Revolutionizing Training with AI Avatars
Synthesia’s core offering enables enterprises to create interactive training videos featuring AI avatars. Their latest innovation, Roleplay Sessions, allows employees to rehearse scenarios such as sales pitches with AI-powered avatars that respond in real time and provide feedback by scoring their answers. This immersive approach transforms traditional training methods by offering scalable, personalized coaching experiences.
At the office opening, I was invited to create my own digital twin—a request I enthusiastically accepted. After all, who wouldn’t want a digital version of themselves, especially when feeling confident in their outfit and appearance for the day?
From Indifference to Intrigue: My Journey with Digital Avatars
Before meeting my digital twin, I was somewhat indifferent toward avatars, though I acknowledged their inevitable integration into everyday online interactions. Many users on platforms like Instagram already employ avatars resembling themselves to generate social media content. This growing trend perhaps explains why I now embrace my digital twin without hesitation. Notably, this marks Synthesia’s first creation of a digital avatar for a journalist—or anyone outside of Alexandru Voica himself.
My avatar is trained specifically on my investigative story about why venture-backed startups commit more fraud than non-VC-backed startups. It is designed to only answer questions related to this research, ensuring focused and accurate engagement.
To interact with the avatar, simply press “start in a new window.” It can answer questions such as:
- Why did I decide to write this story?
- What is the research paper about?
- What did the researchers find?
Behind the Scenes: Creating My Digital Twin
The process to create my avatar took place in a mini film studio inside Synthesia’s office. They captured numerous photos and recorded a two-minute voice sample, all with my full consent. This session resulted in “digital Dom”—a personal avatar that reads any script I provide, available with or without glasses, and two interactive avatars capable of listening and responding, also offered in both versions.
The interactive avatar was trained on my article using a sophisticated tech stack combining voice-to-text, language understanding, text-to-voice, and video animation models. Synthesia’s proprietary video and voice models power the avatar, though customers can select alternatives from leading AI labs such as Cartesia, ElevenLabs, Google, or OpenAI. Enterprises also have the flexibility to host avatars on their preferred cloud infrastructure or opt for Synthesia’s hosting services.
To elaborate, the voice-to-text model transcribes spoken input into text. The agentic language model comprehends and processes this text, generating appropriate responses. These responses are converted back into speech by the text-to-voice model, while the video model animates the avatar’s expressions and lip movements in real time. This seamless integration creates an engaging, lifelike interactive experience.
Synthesia’s Suite of AI Avatar Products
Synthesia offers three primary products:
- A traditional video-creation platform where users input scripts and avatars deliver them as polished video presentations.
- “Sessions,” an agentic platform enabling interactive experiences such as surveys and roleplay scenarios with AI avatars.
- An API platform allowing developers to combine Synthesia’s video and voice models with other technologies to build custom interactive avatars and applications.
Crafting my avatars took only a few days. I initially tested the personal avatar by typing a generic script about the arrival of fall in New York, my favorite season. The AI voice was impressively accurate, notably avoiding the hoarseness present in my original recording.
Reactions from friends ranged from fascination to mild creepiness. The interactive avatar—deterministic and restricted to responding only within its training scope—consistently redirected off-topic questions back to the venture fraud story. Some found the voice less similar to mine compared to the personal avatar, but it remained unsettlingly close. My mother, impressed by the technology, humorously attempted to stump the avatar with personal questions, only to be redirected each time. “I don’t remember giving birth to two of you,” she joked.
The Future of Journalism and Digital Twins
This experience has sparked reflection on the evolving landscape of journalism. Would audiences accept news presented by digital avatars? Some investors immediately dismissed the idea, especially amid current pushback against undisclosed AI-generated content on platforms like Instagram and efforts to limit AI content on sites like Pinterest due to quality concerns. However, others remain uncertain—pondering whether avatars might augment or even replace journalists, or if CEOs would prefer interacting with AI journalist avatars over humans.
Despite these possibilities, the cornerstone of journalism—trust—feels inherently human and unlikely to be fully delegated to AI. My passion lies in connecting with people, conducting research, and crafting stories, elements that build credibility and rapport beyond what an AI can replicate.
Broader Implications of Digital Clones
Beyond journalism, the notion of cloning oneself digitally holds appeal for many professionals. Imagine never needing to catch up after vacations because your digital twin can handle inquiries and maintain presence. The corporate world’s adoption of avatars remains to be seen, but personally, I harbor mixed feelings.
After the initial novelty faded, I found myself observing my avatar intently, waiting—unsure for what exactly. Perhaps for a blink, a smile, or a sign that it truly “knows.” Yet, as my digital twins are deterministic, they will never exhibit spontaneous behaviors. This raises concerns about potential psychological effects if avatars were powered by nondeterministic chatbots capable of freeform dialogue.
I shared with an investor that my generation, Gen Z, may struggle to fully embrace such avatars because they evoke science fiction realities long imagined on screen. Nonetheless, I find digital avatars less unsettling than humanoid robots; at least with avatars, I can always log off when things feel strange.
Until broader adoption unfolds, I invite you to experience my personal avatar, who offers a rundown of the top stories on our site this week!
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Source: Here
