AI The New F-Word of Work: Productivity or Chaos?

Our revolution in artificial intelligence has led to an unexpected result—while AI tools promise faster and more efficient work, the “words describing” the modern office have changed profoundly. It’s not just about technical capabilities; it’s about an entirely new work culture that includes new types of anxiety and stress.

AI programming tools like Claude Code and Opus are no longer just helpful—they have become expected tools that measure each employee’s success and capability. But the real story is more complicated: while top executives experience genuine productivity gains, the majority of the workforce relies on less than 2 hours of time saved per week, or none at all.

How Executives Became ‘Early Adopters’ of AI Coding

The AI programming revolution started from the top. Intuit’s Chief Technology Officer, Alex Balazs, shared his recent morning routine: he begins coding at 5 AM, using AI agents to generate solutions he hasn’t learned in years. “It feels like my brain has changed,” he said, referring to the moment he tried Anthropic’s Opus 4.5 and saw how it solved engineering challenges in 20 minutes.

This enthusiasm from leadership sent a powerful message throughout the organization: AI is not optional; it is essential. Alex Salazar, CEO of Arcade.dev, directly monitors how much his team uses Claude Code, ending meetings with the statement, “You’re not engaging enough yet.” This approach was effective—the company’s AI tool usage increased tenfold after this initiative.

Product managers and designers at Intuit are now adopting “vibe coding”—an approach where they communicate with AI to create working prototypes they can show to engineers. On the surface, it feels empowering. In reality, it has added a new layer of expectation.

The New Metric: From Output to ‘Number of Interactions’

The most alarming development is how companies are changing the way they measure employee performance. At DocuSketch, Vice President of Products Andrew Wirick shared their new metrics: it’s no longer just about finished code, but about the “number of interactions” with AI tools each day.

“The higher the number, the higher the productivity,” is the implicit assumption. Claude Code itself provides weekly reports to each engineer, showing where they get stuck in unproductive conversations with AI.

The problem? These “words describing” metrics create psychological pressure that goes unnoticed. Wirick admits he experiences a kind of compulsion: “I feel like I need to have more interactions each day, and even before sleeping, I think about how to generate more prompts.”

This measurement system set a precedent—productivity is no longer measured by results but by activity.

AI Fatigue: The Unspoken Pressure on Engineers

A UC Berkeley study of an organization with 200 employees revealed a paradox: even though AI handles a large portion of technical work, working hours did not decrease—in fact, they increased.

The new experience words that emerged include: “AI fatigue,” “adaptation anxiety,” “perpetual obsolescence fear.” Engineers constantly worry about missing the next breakthrough, and each breakthrough feels just one prompt away.

Data starkly illustrates:

• 40% of C-level executives believe they save 8+ hours weekly with AI
• 67% of non-management employees report saving less than 2 hours, or none at all
• This gap is not accidental—it’s structural

The Paradox of Productivity: More Work, Less Meaning

Berkeley scholars call this phenomenon “task expansion”—when non-technical collaborators start using AI, engineers become responsible for reviewing and maintaining half-finished prototypes created by marketing teams and product managers.

The result is a large dumping ground of what is called “busyware”: minor website tweaks with no user engagement, custom dashboards for a single user, abandoned prototypes. Each may have a legitimate reason for its creation, but most end up as technical debt.

Balazs said engineering productivity increased by 30% based on code velocity, but the real question is no longer “How fast are we making things?” but “What is the quality of what we make? How many of these codes have lasting value?”

These “words describing”—productivity, efficiency, output—have begun to take on new meanings far from their original intent. Efficiency is no longer about doing meaningful work faster. It’s about doing more work, period.

Reflection Point: Productivity Versus Purpose

As we continue navigating the AI revolution, the critical question is not “How do we maximize AI tools?” but “How do we preserve meaningful work amid the obsession with productivity?”

The “words describing” around us—vibe coding, AI fatigue, task expansion, busyware—are symptoms of a deeper cultural shift. Our measures of success have become misaligned with our true values.

The future won’t be seen in faster coding or more interactions. It will be in how we redesign our relationship with work—how we reclaim space for purpose-driven productivity, not just activity-driven appearances of productivity.