I created an interactive digital avatar of myself — and you can talk to it
- ID
- 28900
- Status
- summarized
- Published
- 26 Sep 2026, 10:00 PM
- Fetched
- 26 Sep 2026, 10:14 PM
- Provider
- TechCrunch
- Category
- technology
- Original URL
- https://techcrunch.com/2026/09/26/i-created-an-interactive-digital-avatar-of-myself-and-you-can-talk-to-it/
- Source URL
- https://techcrunch.com/feed/
Summary
- Score
- 3.0
- Created
- 26 Sep 2026, 10:14 PM
- Tags
- Audience
- vibe_coderssaas_startup_founders
What happened
TechCrunch reporter Dominic-Madori Davis was given a Synthesia AI avatar of herself after visiting the company's new New York office; it's the first time Synthesia has built a digital avatar for a journalist outside of its own staffer Alexandru Voica. The avatar is trained on one of her articles — about why venture-backed startups commit more fraud than non-VC-backed ones — and only answers questions on that topic. Synthesia, originally U.K.-based, hit a $4 billion valuation this year, says it passed $100 million ARR last year, and recently launched Roleplay Sessions for employee practice with interactive avatars; competitors named are D-ID, HeyGen, and Colossyan.
Why it matters
This is a first-person product demo written by someone the vendor invited in, so treat it as marketing rather than evaluation — there is no pricing, API detail, accuracy benchmark, or failure mode disclosed, and the article text is cut off mid-sentence before any conclusion. The one concrete thing builders can act on: a 'digital twin' here means an avatar scoped to a single source document, not a general-purpose clone, so if you pitch or evaluate avatar tools for support, sales training, or content, ask what corpus the model is restricted to and how it behaves outside it.
Discussion angle
If a company's own PR avatar answers press questions about its product, and a journalist's avatar answers questions about her article, where does disclosure belong — and would you let a vendor-trained avatar represent your product to customers?