OpenAI uses 800,000 messages to break down career boundaries: 43

OpenAI used 800,000 messages to break down career boundaries: 43.5% of job inquiries have already crossed their workplace. For the first time, OpenAI put the matter of "AI is smoothing the walls between positions" in front of quantifiable data. It reviewed more than 800,000 work-related ChatGPT messages and came up with a dazzling ratio: 43.5% of job-specific queries actually pointed to work in another profession. The company calls this phenomenon "task cross-fertilization" - people no longer just do the things in their name, but use AI to extend into the territory of neighboring jobs. Marketing and engineering tasks intersect most frequently, like two threads most often intertwined. Users began to personally handle the tasks that used to be left to experts: reviewing contracts, analyzing data, and troubleshooting website failures. These tasks that were originally assigned to legal, analyst, and operation and maintenance positions are now digested by a person holding a chatbot. This momentum is particularly fierce in small companies. They often cannot afford a dedicated team of experts, so non-professionals are more willing to use AI to fill roles such as marketing. Without a full-time team to back it up, AI becomes an easy substitute. OpenAI's judgment is that these traces of usage are an early signal that precedes changes in job titles and job descriptions. The reorganization of functions is already happening, but the organizational chart has not had time to catch up. In order to accurately attribute "tasks" to professional portraits, OpenAI borrowed the American occupational database O*NET to map each activity to a standard occupational classification, and deliberately eliminated common actions that everyone uses such as writing, summarizing, and scheduling. When 43.5% of queries are out of bounds, what is really being rewritten may not be a certain job, but the old yardstick of "how many skills a person should know." via AI News (author: AI Base)