{"version":1,"type":"story","url":"https://digestai.news/story/opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird","json":"https://digestai.news/story/opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird.json","markdown":"https://digestai.news/story/opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird.md","slug":"opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird","headline":"Opinion: I built AI clones of my coworkers and they turned out weird","summary":"The author, a WIRED reporter, used Condé Nast’s Gemini platform to create two AI “clones” of editors Brian Barrett and Sophie Kleeman. By feeding Gemini roughly a month’s Slack messages – about 17,000 words for Sophie Bot – and a few months of public material for Brian Bot, the bots generated style guides, editorial dossiers, and draft ideas. The bots could suggest headlines and even write tweets in the editors’ voices, but they also hallucinated facts, repeated the same phrases, and produced output that felt “grating” and “corny.”\n\nColleagues reacted negatively: Brian called the experiment a “nightmare,” and Sophie described her bot as “a bitch.” A friend who calls himself “Chip” claims his own AI assistant, built with Gemini’s Gem feature, makes him more efficient, echoing a Boston Consulting Group paper that found managers caught 18 % fewer errors in work labeled as coming from an AI employee. The piece also cites McKinsey’s claim that its workforce now includes 40,000 humans and 20,000 AI agents, aiming for a 1:1 ratio by year‑end. The author concludes that while the bots were occasionally helpful, they underscored the limits of AI coworkers and the need to retain human judgment.","keyPoints":["The author fed Gemini about 17,000 words of Slack messages to build a Sophie Bot that could draft tweets and replies.","BCG research found managers caught 18 % fewer errors in work labeled as coming from an AI employee.","McKinsey reports its workforce includes 40,000 humans and 20,000 AI agents, targeting a 1:1 ratio by year‑end."],"whyItMatters":"The experiment highlights practical challenges—privacy, hallucinations, and human‑AI friction—that companies must address before AI coworkers become mainstream.","category":{"slug":"agents","name":"Agents & Tools","url":"https://digestai.news/category/agents"},"entities":{"companies":["Condé Nast","Lattice","Boston Consulting Group","McKinsey & Company","Anything","Stable"],"models":["Gemini"],"people":["Brian Barrett","Sophie Kleeman","Sarah Franklin","Julie Bedard","Dhruv Amin","Bob Sternfels"]},"firstPublishedAt":"2026-09-22T10:30:00Z","updatedAt":"2026-09-22T10:30:00Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"Wired AI","title":"I Built AI Clones of My Coworkers. Things Got Weird","url":"https://wired.com/story/ai-coworker-bots-loved-improv-and-interstellar-soundtrack","publishedAt":"2026-09-22T10:30:00Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Opinion: I built AI clones of my coworkers and they turned out weird\", 22 September 2026, https://digestai.news/story/opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird","publisher":"Digest AI","title":"Opinion: I built AI clones of my coworkers and they turned out weird","datePublished":"2026-09-22T10:30:00Z","url":"https://digestai.news/story/opinion-i-built-ai-clones-of-my-coworkers-and-they-turned-out-weird"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}