At Media Party Buenos Aires 2025, GPTZero founder Edward Tian flipped the conference’s dominant narrative on automation. Instead of discussing how journalists can adopt AI, Tian focused on why current AI tools still need to improve and how we can bridge their critical gaps. He highlighted two major flaws in existing technology: “secondhand hallucinations”—where AI search tools inadvertently cite other AI-generated content as factual sources—and a cognitive risk, backed by MIT research, warning that outsourcing entire drafts to AI degrades human critical thinking skills.
To address these limitations, GPTZero is expanding its mission from basic AI detection to fostering human accountability in the writing process. Rather than just flagging synthetic text, the company now uses browser extensions to analyze typing patterns, verifying authentic human creation as it happens. Furthermore, to combat cognitive decline, Tian’s team is building “human-in-the-driver’s-seat” tools that act as granular editing coaches and outline assistants, ensuring the human remains the primary thinker rather than a passive recipient of automated drafts.
Ultimately, this defense of human intellect is crucial not just for writers, but for the survival of AI itself. Tian revealed a growing paradox: as the web is flooded with synthetic text, AI companies are using GPTZero to filter out AI-generated noise from their training data, as training models on their own outputs severely degrades performance. This feedback loop proves that high-quality human writing and critical thought are far from obsolete; they are the indispensable resource needed to sustain both our minds and the AI ecosystem.
Don’t miss the full presentation! Watch Edward Tian’s Keynote at Media Party Buenos Aires 2025.
Speaker Profile
Edward Tian
Founder, GPTZero
