The "AI Job Apocalypse" is CANCELLED!
TL;DR
Sam Altman and Dario Amodei are changing their tune — Altman said in Sydney that he no longer expects the jobs apocalypse some AI companies predicted, while Amodei now talks about AI as a productivity multiplier rather than a white-collar bloodbath.
Automation is increasing work, not eliminating it — Wes says his own stack of agents for research, sponsorships, and health tracking saves time on tasks but creates more review, decision-making, and follow-up work overall.
The real bottleneck moved from production to judgment — AI can scan headlines, draft copy, code features, and summarize data, but humans still have to frame the task, evaluate results, and decide what matters.
Dan Shipper’s ‘human sandwich’ explains why jobs persist — the human sets direction, the AI collapses the middle of the task, and the human returns to judge and extend the result, which keeps people in the loop even when 90% of a workflow is automated.
Cheap competence triggers a Jevons-paradox effect — because code, writing, design, and support outputs are now far cheaper to generate, companies don’t stop working; they produce far more of everything, which raises demand for expert taste and oversight.
The bigger disruption may hit companies before workers — firms whose moat depended on scarce, expensive competence may get crushed, while AI-native companies with better data, workflows, and human supervision could capture outsized market share.
The Breakdown
The AI job apocalypse may have been overstated: even after automating huge chunks of his own work, Wes Roth says he’s working more, not less, because AI shifts the bottleneck from producing tasks to judging, steering, and owning the output. He argues Sam Altman and Dario Amodei aren’t just doing IPO spin when they soften their warnings — they may be reacting to the same reality many heavy AI users are now seeing.
Was This Useful?
Share
Keep Reading
Make Alcreon Yours
Tune your feedFive quick questions, and the feed ranks what matters to you first.Or just get notified
The weekly Echo. Signal worth keeping in your inbox.
Every new piece, announced on X.
Read Next
See all
Playbook
Tasteful Skills
“Tasteful Skills” argues that the best agent skills are not documentation or best-practice lists.

Playbook
The Art of Tasteful Prompting
Learn how tasteful prompting helps you move beyond generic AI output by shaping context, style, and judgment from the start.

Playbook
The Codex /goal Playbook
OpenAI shipped /goal for the Codex CLI. It turns a prompt into a persisted, self-continuing contract.