Essays & Research

• AI does not widen inequality only through automation, but through unevenly usable access.
• This essay argues that guided onboarding is the missing interface layer between technical availability and real public capability.

Anthropic’s Responsible Scaling Policy matters not only for labs, but for the public-facing layer where most people first meet AI. This essay argues that onboarding design is a safeguarding issue: when capability scales faster than understanding, predictable human failures become part of the risk surface.

• AI has collapsed the cost of producing paper-shaped output, while the cost of deciding what is true has not.
• Nari can enforce verification-first workflows: evidence before prose, provenance by default, and “verification gates” users cannot skip.

• The near-term shock is repricing white-collar work: fewer entry roles, thinner ladders, and faster substitution of routine cognitive tasks.
• Nari helps ordinary users become AI-complementary faster by removing the blank-box barrier and replacing it with guided starting points.

• Teen AI use is already normal; shared understanding between teenagers, parents, and schools is not.
• Nari offers a guided layer that makes AI behaviour more visible, more discussable, and easier to supervise across homes and classrooms.

A recent MyNari event focused on practical family use, guided exposure, and real-world conversation around AI. Use the lead image to open the full gallery.

This page is designed to grow. New essays, event reports, talks, field notes, and research items can be added as further cards without changing the overall structure.