Isabella Calmet's Blog

The Creative Eviction: How AI Forces Us to Reclaim Our Expertise

March 4, 2026

Losing a job feels, almost always, like a blur in one’s identity. For decades, we have defined ourselves by what we do physically: “I write code,” “I analyze X-rays,” or “I draft contracts.” But we are entering an era where the task is no longer the destination, but the starting point[cite: 47, 48, 49].

As I have suggested in previous writings, professional identity is in crisis—but it is not a crisis of extinction; it is one of metamorfosis[cite: 50, 51]. To illustrate this evolution, I want to return to two areas I have explored before: programming and medicine[cite: 52]. Not because the impact is exclusive to them—AI is reconfiguring every corner of the labor market—but because they serve as perfect laboratories to observe how the "loss" of a technical function opens the door to a new form of human value[cite: 53].

The End of the "Code-Monkey" and the "Walking Encyclopedia"

In the world of software, the traditional programmer—the one who spent hours fighting syntax or writing repetitive functions—is seeing their role vanish[cite: 54, 55]. Tools like GitHub Copilot or advanced reasoning models no longer just suggest code; they write and optimize it[cite: 56]. If your value resided in being a “syntax dictionary,” AI has automated your primary asset[cite: 57].

Something similar is happening in medicine. For centuries, a doctor’s prestige was based on their storage capacity: being a walking encyclopedia of symptoms and treatments[cite: 58]. Today, Deep Learning algorithms can process trillions of genomic data points or detect a tumor with a precision that defies the most trained human eye[cite: 59]. Does this mean the end of these professions? On the contrary: it is their liberation[cite: 60].

David Autor’s Thesis: Automation vs. Augmentation

This is where the work of MIT economist David Autor lights the way[cite: 61, 62]. Autor (2015) distinguishes between automation (replacing a human in a task) and augmentation (giving the human new capabilities)[cite: 63]. In his most recent research, Autor (2024) argues that AI is actually “democratizing expertise”[cite: 64].

In medicine, when AI takes over technical diagnosis, it isn't replacing the doctor; it is freeing the professional from the cognitive burden of data retrieval so they can exercise their true competitive advantage: clinical judgment[cite: 65, 66]. The physician evolves from a "data searcher" into an "architect of well-being"[cite: 68].

In programming, the shift is just as profound. The loss of the operational task of writing lines of code is the preface for the programmer to become a "Digital Orchestrator"[cite: 69]. As Brynjolfsson, Li, and Raymond (2023) have shown, generative AI allows workers to move past the "how" and focus on the "what"[cite: 70]. The programmer's expertise is no longer in their fingers, but in their strategic vision[cite: 71].

Expertise as a Compass

Evolution after loss consists of understanding that expertise has shifted[cite: 73]. It is no longer about “knowing how to do,” but about “knowing how to judge”[cite: 74]. As Erik Brynjolfsson and Andrew McAfee (2014) argued, AI is a "General Purpose Technology"[cite: 75]. Just as electricity didn't eliminate factories but forced them to reinvent their design, AI is "evicting" us from routine tasks so we can occupy the higher floors of thought: strategy, synthesis, and empathy[cite: 76].

Overcoming the loss of a job to AI requires a change in narrative[cite: 77]. We aren't being robbed of work; we are being given back the time to do the work a machine will never understand: the work of giving purpose to technique[cite: 78]. This concept of human-centric purpose is something I saw echoed on the Course Dashboard in a recent post by [Insert Classmate Name Here], who discussed the emotional weight of creativity that algorithms can't replicate.

The Paradox of Expertise

However, this evolution is not without its dangers[cite: 79]. There is a crack in the road toward this new expertise: How will we learn to be "orchestrators" if we never learn to play the instruments? [cite: 80] If a young programmer never struggles with basic logic, or if a medical resident never trains their eye on simple cases, we risk creating a generation of "supervisors" who don’t know what they are supervising[cite: 81].

Evolving with AI does not mean delegating our learning; it means using the recovered time to deepen our understanding of the fundamentals[cite: 82]. Human judgment is only valuable if it is grounded in real comprehension, not in blind faith in the algorithm[cite: 83]. True evolution, then, is not just learning to use the tool, but ensuring we don’t lose the "grit" that makes us masters[cite: 84].

References