Artificial intelligence (AI) is rapidly transforming work by automating and redesigning many routine cognitive tasks, including summarization, drafting, coding, scheduling, and basic analysis. These tasks are increasingly converted into low-cost software actions that benefit firms and reduce human workload. As most companies adopt AI across their services, this paper seeks to predict which occupations will be most affected by AI by 2036 using a task-based mathematical model grounded in linear algebra. Jobs are modeled as vectors of task time shares, while AI progress is captured through time-dependent capability and adoption functions. We introduce an AI Exposure Index and a Replacement Pressure Index to distinguish between task augmentation and job substitution. Sensitivity analyses examine outcomes under varying adoption speeds, capability growth rates, and substitution assumptions.
