TechCrunch writer builds interactive AI avatar of themselves

By Splitfeed EditorialLast updated Sep 26, 2026, 18:54 UTC

On September 26, 2026, a TechCrunch writer published a first-person account of building an interactive AI avatar of themselves and making it publicly accessible to readers. The avatar was trained specifically to discuss venture fraud, a narrow and substantive topic, rather than serving as a general-purpose replica. The piece is both a demonstration and a reflection, with the author describing mixed feelings about the experience of creating AI clones of real people.

The story sits at the intersection of two trends that have been accelerating quietly: the commodification of personal AI avatars, and the ethical unease that even enthusiastic early adopters report once they have gone through the process. What makes this more than a novelty piece is that the avatar is live and conversational, meaning readers can interact with something that presents itself as a version of a named journalist. That raises immediate questions about accuracy, misrepresentation, and what happens when a digital clone says something its human original would not endorse. The fact that the author trained it on a specific subject like venture fraud, rather than leaving it open-ended, suggests an attempt to manage those risks, though the limits of that approach are not tested in the available account.

No reactions from the Pro-AI, Anti-AI, or Middle Ground camps have been published yet on this story. Typically, the Pro-AI camp would point to this as a creative and practical expansion of how journalists and experts can scale their knowledge and reach audiences beyond what a single person can manage. The Anti-AI camp would likely focus on the consent and identity risks embedded in cloning a real person's voice and persona, even when that person is the one doing the cloning, arguing it normalises a practice that becomes far more dangerous in other hands. The Middle Ground camp would probably acknowledge both the novelty and the author's own ambivalence as evidence that guardrails and clearer norms are needed before the practice becomes widespread.

The more revealing test will come from reader interactions with the avatar itself. Whether the clone stays reliably within its intended scope on venture fraud, and whether the author publishes a follow-up assessing how it performed, would give a clearer picture of where the real risks and real value actually land.