A long beam on a small fulcrum: a tiny sphere on the long arm lifts a heavy plaster block on the short arm.

1of 3

Hiring in the AI era: find who leverages most

Daniel Dippold’s talent thesis raises a hiring question, not a settled forecast. Test where AI improves useful output before changing the team.

Distribute & Defend
DistributionAI & Automation

Adapted by the EWOR Team from Daniel Dippold’s article (opens in a new tab)2 min read

Start with the work that could change

Daniel Dippold, CEO and co-founder of EWOR, argues that AI could change the distribution of knowledge-work output: people who are already unusually effective might become more productive still. He proposes a shift from a roughly Gaussian distribution towards a power law, with output concentrated among a smaller group. That forecast is unproven; do not build a staffing plan on it.

A more useful question for your own team is: where does a capable person currently lose time to work that tools might help them do, delegate or check? That question can guide an experiment without requiring a prediction about everyone else’s jobs.

Start with the shape of demand

Daniel distinguishes work with relatively bounded demand from work where increased supply might create or serve more demand. The distinction matters because greater individual capacity need not have the same effect in every business.

If customers need only a limited amount of a service, added capacity may alter staffing or pricing choices. If the team can use that capacity to explore worthwhile new products or serve unmet needs, the result could be expansion instead. Neither outcome follows automatically from adopting AI.

Before revising a hiring plan, identify the constraint in your own company. Is the bottleneck skilled attention, customer demand, quality assurance, distribution or something else? A faster contributor has little economic leverage if the next constraint remains untouched.

Run a bounded talent-leverage test

Choose one important workflow and compare an AI-supported version with the existing approach, working with a contributor who knows the task well. Agree in advance what acceptable quality looks like, who checks the output and what time or effort the workflow consumes.

Compare completed work and review burden, not just the speed of generating a first draft. Include the contributor’s view: a tool that creates rework or obscures judgement may offer less leverage than it appears to.

Use what you learn to decide whether to invest in tools, training, clearer responsibilities or an additional hire. Don’t generalise from one person or task to the entire team.

Keep people decisions separate from the forecast

Even if a workflow test succeeds, it does not prove that other employees are replaceable, that performance can be ranked objectively or that changing pay will deliver a predictable return.

Assess whether the investment improves the work while maintaining quality and meeting people-management standards. Any consequential employment decision requires its own evidence and appropriate human review. Speculation about AI-driven job displacement is no substitute.

Source: EWOR article by Daniel Dippold (opens in a new tab)

No comments yet.

Be the first to comment on this signal.

Get the next signal before everyone else.

Frameworks and contrarian theses for outlier founders, straight to your inbox. No fluff, no recycled advice, weekly.

The EWOR Signals newsletter