Meta exec quits, warns AI will hit 1 in 5 corporate roles

Clara Shih's departure from Meta marks a notable shift in perspective from a top tech executive who once helped build AI products. Her account of seeing AI agents reduce a multi-step product development process to one or two people is a concrete illustration of how AI can transform work. She notes that while optimists argue routine tasks get automated, freeing workers for higher-value work, her experience building Salesforce's Agentforce showed that this was 'primarily not true' for customer support, where staff went from around 9,000 to 5,000.

The implications are significant for entry-level jobs, which Shih herself began eliminating at Meta. She predicts that one in five corporate roles will be especially challenged. Her taxonomy, borrowed from MIT economist David Autor, suggests that the impact varies: some jobs face direct substitution, some benefit from automating routine tasks, and others see an influx of workers due to lower barriers to entry, leading to lower-quality jobs, as with taxi drivers. This framework helps explain why AI's effect is not uniform across the economy.

The New Work Foundation's free resources, like Field Report and Game Plan, aim to help entry-level workers navigate this transition. However, Shih admits that 'no one knows' exactly how this plays out, underscoring the uncertainty. Her focus on the transition, rather than the end state, highlights a concern that the path to an AI-driven economy could be disruptive for many workers, even if the destination is positive.