From Hours Worked to Outcomes: Performance in the AI Era

“The management practice of calculating work hours will gradually be eliminated by AI” is a management argument Raymond Hou made in July 2026. His reasoning is that measuring performance by hours assumes that everyone produces roughly the same value per hour. When AI creates a several-fold gap between people who can use the tools and those who cannot, that assumption no longer holds. Measurement must shift from time invested to outcomes and impact produced. ¹

Starting Point: The Failure of Time Tracking

A specific observation prompted this argument. Raymond’s teammates found that time tracking became harder after adopting AI: tasks were no longer carried out in one continuous block, and several things could advance at once within the same time period. Boundaries between the time spent on individual tasks became blurry. Raymond’s own work was similar: he might write a newsletter while sending ideas for changes to an AI Agent, and more than an hour would pass on the clock even though he had only 15 to 30 minutes of focused effort. He concluded that the problem was not that time-tracking tools were inadequate; the unit of “hours worked” itself no longer corresponded to value.

This aligns with an earlier personal metric: since early 2026, Raymond has stopped measuring his work by the hours he sits at a computer and instead tracks how much AI usage quota he consumes in a week (see Raymond’s AI Double).

A Criterion: Achieve the Same Goal With Fewer Iterations

As evidence, Raymond cites a scoring mechanism he considers well designed. In a vibe-coding competition at a Google Cloud annual event, one scoring criterion awarded a higher score for using fewer tokens. He interprets this to mean that achieving the same goal with fewer correction rounds reflects a stronger understanding of the problem, more accurate instructions, and higher quality standards. In other words, the new measure of performance is not “how long did it take?” but “what verified result was achieved, at what cost, and with how many iterations?”

Implication: Reintegrating Roles

Raymond further derives an organizational change. In the past, divisions of labor existed because individual capacity was limited and people had to specialize in one function. When execution can be delegated to AI, walls between functions no longer need to remain. He gives three examples:

• Customer service and sales: Both deal with the same consumers, but were previously separated due to staffing limits. After integration, a person can handle the part they are good at while AI operates the other part, connecting proactive recommendations, repeat-purchase rates, and user feedback in one workflow.
• Design: The work no longer ends at delivering attractive visuals. It can extend to product promotion, list conversion rates, and newsletter automation, understanding the interaction data, retention, and SEO affected by each image.
• Administrative work: Rather than a cost center seen as “unrelated to making money,” scheduling and office space can be considered in terms of their effect on productivity and retention, becoming part of the overall synergy.

The shared conclusion is that when AI handles execution, people should shift from thinking “what should someone in my function do?” to thinking like an operator: how does this connect to revenue and growth? Raymond imagines a future where “one person plus an AI team is a department,” and agents collaborate across departments with other agents.

Position in Raymond’s Framework

This proposition brings his long-standing resource hierarchy (see Attention > Time > Money) into management: if attention is scarcer than time, using time as the unit of payment and evaluation was already a second-best solution. AI has simply made the gap impossible to ignore. It is also the organizational version of The Super AI Individual: Extending a Working System from People to AI: at the individual level, the concept concerns delegating work to AI; at the organizational level, it concerns how to measure the remaining human work. This is why Raymond sees AI capability as a career lever rather than a tool skill: in an outcome-priced environment, the difference between people who can amplify results and those who cannot will be directly reflected in their value.

Source

Footnotes

1. Raymond Hou, “Will the Management Practice of Calculating Work Hours Gradually Be Eliminated by AI?” public Facebook post, 2026-07-11. View original ↩