Visualizing Careers in Flux: Making Sense of AI's Disruption of Work and College Pathways—Through an Informed, Skeptical Lens

This study investigates how AI is reshaping early-career choices using a 120-response mixed-methods survey, qualitative testimony, and task-level datasets linking worker-desired offloads to expert-rated feasibility. Findings indicate that adoption is real but uneven: frequent workplace use remains a minority, while experimentation and light, task-specific use are common. Training exposure, weighted by usage intensity (daily/few-times-a-week = 1; occasional = 0.5; rare/never = 0), shows self-taught/basic exposure predominates, with a significant proportion reporting no formal training. Among active users, perceived necessity correlates with perceived productivity gains, consistent with utility-driven adoption. Offloads concentrate on writing/drafting, coding/development, and data/BI tasks. Organizational stance is concentrated in experimenting and active integration, with fewer respondents reporting no consideration. Qualitative themes highlight benefits alongside concerns about privacy/security, bias/fairness, over-reliance/loss of autonomy, and job displacement. A worker desire vs. expert feasibility gap is observed at the task level; the miscalibration computation is detailed in Methods/Appendix. Overall, the evidence supports a staged transition in which tasks change first and jobs re-bundle subsequently, providing actionable guidance for program selection and early-career planning.