The panic and reality

Public conversation oscillates between "AI will eliminate 50% of jobs" and "AI will create more jobs than it destroys". Reality, according to Goldman Sachs, WEF and BCG, is more nuanced: ~16K net jobs/month eliminated in US, ~170M new roles by 2030 globally.

Roles that change

Most cases aren't elimination — they're transformation. Software developer: still exists but uses Cursor/Claude, 3-5× more productive. Marketing analyst: uses AI for research and drafts, focuses more on strategy. Lawyer: uses AI for document review, focuses more on negotiation.

Roles that disappear

The most exposed: basic data entry, simple translation, L1 customer service, simple content moderation, basic transcription. The pattern: high-volume tasks with clear criteria and limited judgment.

Roles that emerge

AI/ML Engineers: most in-demand category. Prompt Engineers: from $200K+. AI Trust & Safety: compliance, ethics, alignment. AI Agent Owners: responsible for autonomous systems. Skilled trades: electricians, plumbers, welders — AI doesn't compete (yet) with complex physical work.

The Gen Z gap

The most underestimated impact: entry-level hiring fell 25% in top 15 tech 2023-2024. AI does the tasks juniors previously did. Yale calls it "the real job destruction is happening before careers can start".

The gender bias

79% of working women in US have jobs with high automation risk vs 58% of men. Reason: concentration in administrative, clerical, customer service roles. Required: specific reskilling programs for this population.

How to adapt

For individuals: (1) AI fluency in your current role. (2) Identify high-judgment vs high-repetition in your work. (3) Build portfolio of "what you did with AI". For companies: (1) Don't lay off in anticipation. (2) Invest in upskilling. (3) Redesign roles to combine human + AI.

Conclusion

The transition is real, painful for some, opportunistic for others. The 2030 net balance is positive but the cost is distributed unevenly. The right policy response is: serious reskilling, support for transitions, redesign of education from K-12 to university. Both panic and triumphant techno-optimism are wrong responses.