When Allie Abusamra graduated from a small private college in New England with a degree in English, nobody handed her a golden ticket to Silicon Valley. There was no obvious pipeline from literary analysis to artificial intelligence. No recruiter waiting at the career fair with a Google badge. She found her way there anyway — and what happened next tells us something significant about where the technology industry is heading and who it actually needs.
Abusamra’s story, first reported by Business Insider, traces an unconventional path from humanities coursework to a position at the center of Google’s AI efforts. She didn’t learn to code. She didn’t pivot to a computer science boot camp. Instead, she brought to the table exactly what her English degree trained her to do: read carefully, write precisely, and understand how language actually works.
That turned out to be exactly what Google needed.
At Google, Abusamra works on AI-related projects where her responsibilities involve crafting, editing, and evaluating the language that large language models produce. She assesses whether AI-generated text reads naturally, whether it conveys meaning accurately, whether it captures the right tone. This is not a peripheral task. It sits at the core of what makes products like Gemini useful or useless to hundreds of millions of users. And it requires a kind of expertise that no amount of Python proficiency can replicate — the ability to parse ambiguity, detect subtle failures in reasoning, and understand what humans actually mean when they string words together.
Her English degree, as she told Business Insider, became an edge rather than a liability. The close reading she’d practiced on novels and poetry translated directly into the close reading of AI outputs. She was trained to notice when something was off — a misplaced emphasis, a logical gap disguised by fluent prose, a confident assertion that didn’t hold up under scrutiny. Large language models are exceptionally good at sounding right. Identifying when they’re wrong requires a different skill set entirely.
This matters because the AI industry has a language problem it can’t engineer away.
The current generation of AI systems — GPT-4o, Gemini, Claude, Llama — are fundamentally language machines. They predict the next token in a sequence. They generate text. They interpret prompts. Every interaction a user has with these systems is mediated by language, and the quality of that mediation determines whether the product feels intelligent or merely verbose. Companies have spent billions on compute, on training data, on architectural innovations. But the final layer — the one the user actually experiences — depends on people who understand how language works at a granular, human level.
Google is not the only company recognizing this. Across the industry, demand has been rising for what some firms call “language specialists,” “AI trainers,” or “prompt engineers” — roles that often draw on humanities backgrounds rather than STEM credentials. Anthropic, the maker of Claude, has hired philosophers. OpenAI has brought on writers and editors. Scale AI built much of its business on human evaluation of AI outputs, employing thousands of contractors with diverse academic backgrounds to assess model quality.
But Abusamra’s role at Google suggests something more substantial than a trend in contractor hiring. She isn’t evaluating outputs in a gig-economy queue. She’s embedded in the company, shaping how AI products communicate. That distinction matters. It signals that the value of humanistic expertise isn’t limited to quality assurance — it extends into product design, user experience, and the fundamental question of what these systems should sound like when they talk to people.
The tech industry’s relationship with humanities graduates has historically been, to put it charitably, ambivalent. For years, the dominant narrative was clear: STEM degrees led to high-paying tech jobs; liberal arts degrees led to barista positions and regret. This narrative was never entirely accurate — plenty of tech executives studied philosophy or history — but it shaped hiring practices, university enrollment patterns, and parental anxiety in measurable ways. Computer science enrollment at U.S. universities roughly tripled between 2013 and 2022, according to the Computing Research Association. English departments, meanwhile, saw enrollment declines so steep that some universities began cutting faculty lines and merging programs.
The irony is sharp. Just as students fled the humanities, the tech industry built products that desperately need humanistic expertise to function well.
Consider the specific challenges that large language models present. They hallucinate — generating plausible-sounding information that is factually wrong. They exhibit biases embedded in their training data. They struggle with nuance, context, and the kind of pragmatic reasoning that humans perform effortlessly in conversation. They can write a sonnet but may not understand why a particular metaphor is inappropriate in a medical context. Solving these problems requires more than better algorithms. It requires people who can identify failures that are linguistic and conceptual rather than computational.
Abusamra’s experience at Google illustrates this concretely. As she described to Business Insider, her work involves evaluating AI text for qualities that are inherently subjective but critically important: clarity, coherence, appropriateness, accuracy of tone. These are judgment calls, not binary assessments. They require the kind of interpretive sophistication that English programs have been teaching for centuries — the ability to hold multiple readings of a text in mind simultaneously, to consider audience, to weigh connotation against denotation.
None of this is to suggest that English majors should replace software engineers. The point is narrower and more practical: building AI products that work well for humans requires human expertise that the tech industry has historically undervalued. And the companies that figure this out fastest will have a meaningful advantage.
Google appears to be figuring it out. The company has been expanding its investment in what it calls “responsible AI” and human evaluation processes. Its approach to Gemini, the multimodal AI model that competes with OpenAI’s GPT-4o, involves extensive human review of outputs across languages, domains, and use cases. People like Abusamra are part of that apparatus — not as afterthoughts, but as essential contributors to product quality.
The broader labor market is starting to reflect this shift, if unevenly. A recent analysis from LinkedIn showed growing demand for roles combining communication skills with AI fluency. Prompt engineering, which barely existed as a job category three years ago, now appears in thousands of postings across industries from healthcare to finance. And while many of these roles don’t explicitly require humanities degrees, the skill sets they demand — analytical writing, critical reading, rhetorical awareness — map closely onto traditional liberal arts training.
There’s a deeper philosophical question here, too. As AI systems become more capable of generating text, the ability to critically evaluate that text becomes more valuable, not less. When anyone can produce a 2,000-word report in seconds, the scarce resource isn’t production — it’s judgment. Knowing whether the report is accurate, well-reasoned, appropriately framed, and genuinely useful to its intended audience. That’s editorial thinking. It’s the thing English departments have been teaching all along.
So what does Abusamra’s story actually tell us?
It tells us that the skills gap in AI isn’t only technical. It tells us that companies building products made of language need people who understand language — not just statistically, but humanistically. It tells us that the decades-long devaluation of humanities education may have been not just culturally impoverishing but strategically foolish. And it tells us that if you’re a hiring manager at an AI company, the next resume you should pay attention to might have “English” where you expected “Computer Science.”
Abusamra didn’t plan to end up at Google working on AI. Few English majors in 2020 would have imagined such a path existed. But the path exists now, and it’s widening. The question is whether the industry — and the universities feeding into it — will recognize this quickly enough to act on it. The students who chose English despite the cultural pressure to study engineering may turn out to have made a better bet than anyone realized. Including, perhaps, themselves.
WebProNews is an iEntry Publication