“Hours worked ≅ performance” assumes that everyone produces roughly the same value per hour. AI breaks that assumption; measurement must shift to outcomes and impact.

Raymond’s explanation

Time tracking has become harder in the AI era: tasks are no longer a single stretch of continuous effort, but several things progressing within the same block of time. An hour passes on the clock, but actual focused work may have lasted only fifteen minutes. The problem is not the tracking tool; it is that the unit “hours worked” no longer corresponds to value.

The new criterion is what verified result was achieved, with how much cost and how many iterations. Reaching the same goal with fewer correction rounds means a stronger understanding of the problem, more accurate instructions, and higher quality standards.

When to use it

  • Team management and performance design: shift from “how long did it take?” to “what was produced and what was solved?”
  • Self-measurement: since early 2026, Raymond has used weekly AI quota consumption instead of hours spent sitting at a computer.
  • Organizational design projections: when execution is handed to AI, walls between functions (customer service / sales, design / marketing, administration / operations) lose their reason to exist; “one person + an AI team is one department.”

Supporting evidence

  • A team member’s firsthand test: time tracking became harder after AI was introduced.
  • A Google Cloud annual-conference vibe-coding competition: the fewer tokens used, the higher the score.

Counterexamples and boundaries

  • This does not mean ignoring effort: learning, relationship building, and developing taste are “slow work” that should not be priced by immediate outcomes.
  • Outcome-based measurement requires verifiable acceptance criteria; otherwise it regresses to “looks like output.” See the Verify field in Externalizing Judgment.

Origin

Facebook post “Will Methods for Calculating Work Hours Gradually Be Replaced by AI?” (2026-07-11); full argument in the raymond-wiki From Hours Worked to Outcomes entry.

Where it has been discussed

  • From Hours Worked to Outcomes: full reasoning and three examples of combining functions
  • Super AI Individual: the personal version (how to hand work off) and the organizational version (how to measure it afterward)
  • Raymond’s Newsletter #184: the section “Methods for Calculating Work Hours Will Gradually Be Replaced by AI”

Attention > Time > Money, Super AI Individual, Externalizing Judgment, Time Management, Business Thinking