A look at South Korean conglomerate Naver's experiments with a fleet of ~100 office robots, including efforts to make employees comfortable with their presence
without making employees uncomfortable or violating their privacy. @daiwaka and I spent time at Naver's robot-friendly building to find out how the robots work. https://www.nytimes.com/... Matthew Boyle / @bizboyle : “Putting the robot at the back of the elevator made humans uncomfortable...” https://www.nytimes.com/... Daisuke Wakabayashi / @daiwaka : Robots are frequently used in factories, and in hospitality but they're still rare in the white-collar workplaces of offices and cubicles. Korea's Naver is working on how to turn robots into polite and useful office mates. w/ @johnjyoon https://www.nytimes.com/...
Context & Ripple Effects
When reporters embedded with Naver's ~100-robot fleet in late 2022, the hard problem wasn't locomotion or delivery — it was etiquette: even something as small as where a robot stands in an elevator changed how humans felt about sharing the building. That acceptance-first framing set up what came after: Naver turned the experiment into commercially available Rookie helper robots within about a year, then into the 1784 tower, where thousands of employees now work alongside 120+ robots daily.
First-order effects
- Naver's own employees become the test population: their discomfort signals — like robots crowding elevators — directly shape the hardware placement, routing, and privacy rules of the fleet they work beside.
- The office itself becomes a controlled lab: every awkward encounter feeds design decisions that determine whether the robots stay an internal curiosity or ship as a product.
Second-order effects
- Once the etiquette playbook proved out, Naver's move to sell Rookie robots to other companies turned workplace acceptance from an HR question into a sales prerequisite — buyers will judge the robots by how their own staff react to them.
- Rivals eyeing service robotics inherit a benchmark: a large conglomerate demonstrating that white-collar deployment requires comfort engineering, not just capability, raises the bar for anyone pitching office robots without it.
Third-order effects
- If the pattern holds, offices follow factories and hospitality as a standard robotics proving ground — with the twist that the 'training data' includes human tolerance, making employee consent and privacy norms part of the product spec rather than an afterthought.
- Deployment economics may increasingly depend on the human layer around the machines — the same dynamic visible in remotely piloted systems like Telexistence's store-restocking robots — pushing robot vendors to compete on integration and acceptance as much as autonomy.
The trend: White-collar workplaces are shifting from robot-free zones to deliberate testbeds where human acceptance and privacy engineering gate the path from internal pilot to commercial product.