The zero value of average work

The economics of knowledge work are evolving rapidly due to advancements in AI technology. Prior to AI, it was widely recognized that producing high-quality work at scale necessitated time, specialized skills, and organizational infrastructure. This resulted in meaningful costs associated with activities such as writing articles or developing functional software.

However, the emergence of LLMs (Large Language Models) has significantly reduced these costs to virtually zero. Consequently, AI-generated output has become the new standard, and "competent" execution is now readily available at minimal cost. This shift in value dynamics is not unique to AI; similar transformations have occurred in the past, such as with cloud computing and GPS technology.

For CEOs, this presents a critical question: when knowledge work becomes abundant and inexpensive, where does the value lie? One industry that offers insights into this question is search marketing. Historically, search marketing required expertise across various functions, including content production, keyword research, technical recommendations, reporting, and analysis.

These capabilities necessitated specialized personnel and processes, along with advanced technology. Consistent and efficient execution of these functions provided a competitive advantage. However, AI has lowered the cost of many of these functions while concurrently generating a surge of new content, dashboards, trackers, and recommendations.

The competitive advantage has shifted, but the ability to discern which work generates meaningful outcomes remains a human skill. AI has the potential to perform every step of the content workflow, from research and writing to quality review and publishing. However, human intervention is still essential because expert decisions at each step impact the final outcome.

People must evaluate sources, provide industry context, and determine whether the activity aligns with the client's goals. This becomes particularly crucial during optimization. Businesses have historically optimized measurable proxies like rankings, clicks, backlinks, traffic, and content volume because these metrics correlated with business outcomes.

However, the relationship between these proxies and business impact is evolving. A ranking may lead to reduced traffic as consumers increasingly obtain answers directly from search engines and AI platforms. Traffic may not accurately reflect whether a brand is visible throughout a customer's research process. Content volume can increase without enhancing authority, awareness, or demand.

Leaders must continually reassess whether the SEO metrics their organizations prioritize still correlate with desired outcomes. There is a risk of optimizing the wrong thing, as lower execution costs amplify the impact of flawed decisions. When organizations can produce content, software, analyses, and campaigns swiftly and inexpensively, erroneous assumptions can proliferate before being challenged.

The bottleneck shifts from generating work to deciding what work merits production. This is where experienced practitioners can add significant value. Expertise encompasses the accumulated context that enables individuals to recognize weak proxies, question assumptions, connect information across different business areas, and identify when an efficient course of action leads the organization astray.

It also involves knowing when to cease action. In an era where virtually anything can be optimized or automated, deciding what does not merit optimization may become one of the most valuable skills an individual can develop. Organizations should build around judgment, a perspective that should prompt CEOs to reconsider talent differently.

Asking how much of someone's current job AI can perform provides only part of the answer. It is equally important to consider how that person can leverage AI as a force multiplier. Employees who possess deep knowledge of their field have an opportunity to work on a broader range of problems, applying their experience where judgment and context are paramount.

As AI tools continue to improve, many current AI skills will become commonplace, and interfaces will become more user-friendly. Consequently, differentiation will increasingly rely on the expertise, experience, and perspective that individuals bring to the tools. Over the next several years, as AI technology advances, the scarcity of value will likely reside in the decisions people make and the judgments they apply to the tools.