For a generation of white-collar professionals, the primary anxiety of the modern workplace has shifted. It is no longer just the fear of being fired—a sudden, sharp event—but something more corrosive: the fear of becoming irrelevant. This psychological shift has coalesced into a new acronym, FOBO, or the Fear of Becoming Obsolete.
Unlike traditional job insecurity, FOBO is a quiet, persistent dread that one’s hard-won skills are evaporating in real-time. The numbers suggest this is not mere paranoia. According to data from KPMG, four in 10 workers now identify AI-driven job loss as a primary fear, a figure that has nearly doubled in a single year. This anxiety is compounded by a sense of alienation; roughly 63% of employees report that AI is making the workplace feel less human.
The catalyst for this angst is a series of stark predictions from the architects of the technology itself. Dario Amodei, CEO of Anthropic, and Mustafa Suleyman, CEO of Microsoft AI, have both suggested that AI could eliminate up to 50% of entry-level white-collar positions within five years. Even in the halls of government, the alarm is sounding. Senator Mark Warner (D-VA) recently noted that AI leaders are “literally consciously pulling back on their predictions” to avoid triggering immediate economic chaos, while projecting that unemployment for new college graduates could hit 35% within two years.
However, a comprehensive new study from MIT FutureTech suggests that while the direction of this fear is accurate, the timeline may be less apocalyptic than the headlines suggest. The researchers argue that the march of AI is not a series of “crashing waves” that wipe out specific roles overnight, but rather a “rising tide” that lifts—or submerges—capabilities broadly and incrementally across the entire economy.
The ‘Rising Tide’ vs. The ‘Crashing Wave’
The MIT study, titled “Crashing Waves vs. Rising Tides,” represents one of the most rigorous empirical looks at AI’s actual performance in the labor market. Led by researchers Matthias Mertens and Neil Thompson, the team analyzed more than 17,000 evaluations of LLM outputs across 3,000 distinct labor market tasks derived from the U.S. Department of Labor’s O*NET system. They tested over 40 models, including frontier systems like GPT-5, Claude Opus 4.1, and Gemini 2.5 Pro.
The central question for those gripped by the Fear of Becoming Obsolete is simple: Can AI do the job well enough that a manager would accept the output without edits? The answer is increasingly yes. Currently, AI successfully completes between 50% and 75% of text-based labor tasks at a minimally acceptable level. By the third quarter of 2024, frontier models were already hitting a 50% success rate on tasks that typically require a full human workday.
The trajectory is steep but predictable. Failure rates are halving roughly every two to three years, which suggests annual gains of 15 to 16 percentage points in success rates. In an optimistic, upper-bound scenario, MIT researchers project that AI could complete most text-based tasks with an 80% to 95% success rate by 2029.
This “rising tide” framing is critical because it implies visibility. Rather than facing a sudden “discontinuous jump” into obsolescence, workers are likely to see the water rising. This creates a window for strategic adaptation, though the researchers warn that “gradualism is not inherently protective.”
AI Success Rates by Profession
The impact of this tide is not uniform. Some sectors are closer to the inflection point than others, based on the AI’s ability to complete text-based tasks without human intervention.
| Profession/Domain | AI Success Rate |
|---|---|
| Installation, Maintenance, and Repair | 73% |
| Healthcare Practitioners | 66% |
| Business and Financial Operations | 57% |
| Management | 53% |
| Legal Operate | 47% |
The Institutional Adoption Paradox
There is a profound irony at the heart of the current economic moment: while AI’s capabilities are soaring, corporate adoption is lagging. FOBO is not just an individual condition; it is an organizational one. Data from Goldman Sachs economists Sarah Dong and Joseph Briggs, citing Census Bureau data in their March 2026 AI Adoption Tracker, shows that fewer than 19% of U.S. Establishments have actually adopted AI. They project this will only rise to 22.3% over the next six months.
This paralysis is largely a matter of “plumbing.” A survey of 500 business leaders revealed that 83% lack the necessary data infrastructure to fully leverage AI. The technology is ready, but the organizational pipes are not. This gap is leaving workers to navigate their fear in a vacuum; research from the workforce nonprofit JFF indicates that only about one-third of workers feel their employer is providing adequate AI training or reskilling opportunities—a decline of nearly 10 percentage points since 2024.
Despite this institutional lag, the cost of inaction is measurable. Enterprise workers who do integrate AI are recapturing 40 to 60 minutes of productivity per day, according to OpenAI data from December 2025. For a team of 50, this translates to roughly 33 to 50 hours of recovered productivity every single day. Goldman Sachs economists note that while academic studies imply a 23% average uplift in productivity, company anecdotes suggest gains as high as 33%.
The High Cost of Resistance
Inside the firms that are successfully deploying these tools, a new and dangerous divide is emerging. Joe Depa, the global chief innovation officer at EY, describes a generational and psychological sorting process. While junior employees tend to adopt AI tools immediately, senior leaders often lag.
More concerning, however, is the “high-skill resister.” Depa has observed experienced software engineers—individuals who were once “top of their class”—refusing to employ AI because they believe they can perform the task better manually. The result is a productivity collapse relative to their peers. While some “mediocre” workers have used AI to become top performers, the resistant experts have, in some cases, fallen to the bottom of their peer groups.
In these instances, the Fear of Becoming Obsolete becomes a self-fulfilling prophecy. By resisting the tool to protect their sense of mastery, these workers accelerate the very irrelevance they dread. For those whose productivity lags their peers by 10x or 20x, the result is inevitable: they will eventually have to discover a different role.
The bottom line is that AI is not a verdict, but a tool. The transformation of the labor market is likely to be a steady, visible tide rather than a sudden drowning. This means the window to adapt is open, but it is not infinite. The antidote to FOBO is not denial, but the willingness to treat one’s expertise as a foundation for AI augmentation rather than a replacement for it.
The next critical marker for the economy will be the upcoming quarterly labor reports and the 2026 mid-year AI adoption updates from the Census Bureau, which will reveal whether corporate infrastructure is finally catching up to the technology’s potential.
Do you feel the “rising tide” in your industry, or are you seeing “crashing waves”? Share your experience in the comments below.
Disclaimer: This article is for informational purposes only and does not constitute financial or career advisory services.
