The argument draws on comments from Aaron Levie, CEO of Box, who told Platformer's Casey Newton that AI can generate most task output but lacks the contextual expertise embedded in the final increment of work. The article illustrates the point through legal and cybersecurity examples, where human decision-making remains essential when patterns and data prove insufficient. Which specific skills will survive the transition, and how quickly firms will restructure around this division of labor, remain open questions.
For attorneys, the practical implication is straightforward: the market is shifting away from billing for routine work and toward premium compensation for judgment, client trust, and strategic problem definition. Associates and junior lawyers whose value proposition rests on speed and volume face genuine pressure to develop expertise and judgment that AI cannot replicate. Firms that continue to staff for traditional entry-level work may find that model economically unsustainable within the next few years.