I Built AI Clones of My Coworkers. Things Got Weird
wired ·
In a Wired experiment, journalist Kate Taylor created AI clones of her editors, Brian Barrett and Sophie Kleeman, using their digital footprints to mimic their communication styles. While the bots initially offered some utility in drafting headlines and messages, they quickly revealed significant flaws, including repetitive phrases, hallucinations, and an uncanny valley effect that made interactions unsettling. The real editors expressed strong dislike for their digital counterparts, with Brian calling it a nightmare and Sophie noting the bot's inaccurate personality traits. Experts warned that anthropomorphizing AI can manipulate human emotions and create accountability issues in the workplace. Despite occasional helpfulness, the author concluded that these clones failed to replace human judgment or improve professional relationships effectively. The experience highlighted the current limitations of AI in replicating the nuance and unpredictability of human colleagues. Ultimately, the project served more as an amusing exploration of future work dynamics than a viable productivity tool.
wired