New Browser Plugin Adds Typos to Your AI-Generated Emails to Make Them Look Real

Why It Matters

As AI‑crafted communications become increasingly scrutinized, Sinceerly offers a novel way to preserve perceived authenticity, influencing email etiquette and AI‑detection strategies.

Key Takeaways

• Sinceerly injects deliberate typos into AI‑written emails.
• Users can select error severity, from subtle to “CEO” level.
• Creator Ben Horwitz built it using Anthropic’s Claude AI.
• Test with five 500 CEOs produced four brief replies, two with typos.
• Plugin currently suffers bugs, limiting practical use.

Pulse Analysis

The proliferation of large‑language‑model writing assistants has transformed how professionals draft emails, reports, and even academic papers. While these tools boost productivity, their immaculate output often raises red flags among managers and educators who suspect automation. This growing skepticism has sparked a niche demand for solutions that re‑introduce human imperfections, a gap Sinceerly seeks to fill by programmatically inserting misspellings, errant capitalization, and casual sign‑offs.

Sinceerly’s architecture leverages Anthropic’s Claude to reinterpret a polished draft and then apply a configurable error layer. Users can toggle between "Subtle"—minor punctuation slips—and "CEO," which adds overt mistakes and even the classic "sent from my iPhone" footer. In a self‑reported test, Horwitz emailed five Fortune 500 CEOs; four responded, two with typographical errors, suggesting the tactic may increase reply rates by lowering perceived AI usage. However, the plugin’s recent rollout suffers stability issues, with many users encountering error messages that halt its functionality, limiting its immediate utility.

If the technical glitches are resolved, Sinceerly could spawn a new micro‑market for AI‑counterfeit tools that help users navigate the fine line between efficiency and authenticity. Companies may adopt similar solutions to comply with emerging AI‑disclosure regulations, while marketers might exploit the approach to humanize automated outreach. Yet the ethical dimension remains contentious: deliberately degrading content could erode trust if discovered. The conversation around intentional imperfection underscores a broader industry challenge—balancing AI’s convenience with the need for genuine, relatable communication.