One Graduate’s Search Tests the AI Hiring Story

In a February newsletter, Business Insider columnist Alistair Barr argued that people should not panic about AI destroying tech jobs. Kiran Maya Sheikh, a recent University of California, Irvine computer science graduate, wrote to disagree. She said AI was disrupting the job market and that employers were prioritising experienced workers over new graduates. The exchange led to a March 2026 interview, and Barr recounted it in an October 2026 column.

Her case matters because it puts a human story at the centre of a dispute often conducted in abstractions. Is AI narrowing the first rung of a tech career? Or are graduates facing a harsher version of a familiar competition for junior roles, shaped by hiring caution as much as by automation? Sheikh’s experience offers detail. It cannot answer the question on its own.

A promise of stability, then a difficult search

Sheikh told Business Insider Africa in a March 2026 interview that she entered UC Irvine in 2020, enjoyed her first coding class and was drawn to computer science mainly by the prospect of a stable job and enough money to support her family. She also wanted financial independence. The plan was conventional: study the subject, build skills, find a company that was hiring.

She graduated in June 2025 and soon found the market much harder than she had expected. Many classmates appeared not to be getting opportunities, she said. She described a race to use AI to improve and send out CVs. The search felt more intense than she was prepared for.

In her account, the imbalance was visible in the listings. She said she saw more mid-level openings than entry-level ones, and felt she was competing with AI and other graduates for roles that did not yet exist. She had expected AI to be integrated into software engineers’ work. She had not expected it might take over the jobs she was seeking. Looking back, she said she should have researched the job market and predictions about AI more carefully.

She questioned her decision to study computer science but did not abandon the subject. She preferred to specialise and keep up with new information rather than switch discipline, hoping to offer something AI could not.

What one search cannot measure

The difficulty Sheikh describes is real to her, but its causes are not measured. Her account does not apportion the problem between AI, competition from experienced engineers, fewer entry-level listings, hiring freezes, cost-cutting or broader caution. She was reading job portals and observing classmates, not analysing a labour-market dataset. The available reporting does not include application numbers, vacancy counts or a comparison group that would show whether her experience was typical.

That limitation matters because the wider debate often slides from anecdote to trend. Barr’s column is a personal column built around one graduate’s correspondence, not a labour-market study. The March 2026 interview and the October 2026 column were both written by Barr and published by Business Insider properties, so they add detail but are not independent of each other. The appointment itself was reported from Sheikh’s email to Barr and was not independently confirmed.

The offer, and what it qualifies

By June, before the offer, the column reported that Sheikh was still searching while learning new skills, building websites and meeting people. She had volunteered, pursued fellowships and built relationships rather than waiting for a perfect opening, according to the column. She said living at home with family reduced her expenses, and that finding professional communities helped her confidence.

In the October column, she told Barr that she had started her first full-time job as a software engineer at the Virginia Digital Service, described as Virginia’s state government technology modernisation team. She had moved from Los Angeles to Richmond, Virginia. The outcome is good news for one graduate. It qualifies the darker reading of her earlier interview: the entry-level market was not closed to her.

It does not settle the wider question. One reported job offer cannot show that the entry-level market is healthy, that AI has not affected hiring or that her strategy caused the result. Sheikh’s own lessons were that networks matter and that results come from effort combined with luck, with chances improved by doing as many things as possible. That is a reflective account of one search, not a tested formula. The column presents volunteering, fellowships and networking as part of her strategy. It does not show that they produced the offer.

Why the warehouse example matters, and where it stops

Barr’s column reaches for a different case to make a broader point about automation. It reports that an Amazon warehouse in Tracy, California, employed more than 3,000 people, up from roughly 1,000 when it opened in 2023, according to Amazon. Workers still picked, packed and loaded packages, while others learned to supervise and maintain machines. Amazon said its more automated warehouses support roughly 30 per cent more skilled jobs.

Those figures are company claims and were not independently verified. The column also notes that Amazon had explored using robots to slow future hiring growth, as reported by Eugene Kim in May 2025. Both points can be true at once. Automation can change the composition of work and the future trajectory of hiring without immediately eliminating jobs. A site can add workers while needing fewer of them per package, or while planning to add fewer in future.

The warehouse example is a useful warning against equating automation with instant unemployment. It is less useful as a direct test of entry-level software hiring. The tasks, the skills and the labour market are different. Barr’s column also cites his recent reporting that tech openings had jumped that year, partly because of AI-related demand, while tech layoffs remained well below their 2022 peak. Those are the columnist’s characterisations; the column does not reproduce the underlying figures or date ranges.

Two narratives, one missing dataset

The Hack Academy’s earlier commentary on the coding degree took a darker view, arguing that recruitment freezes and generative AI had collided with the junior tasks graduates were trained to do. That was an argument, not an independent measurement, and it does not settle the question either. The disagreement between the optimistic and pessimistic readings is less a direct factual contradiction than a difference of emphasis and evidence. Neither account includes the primary data that would resolve it.

What would that evidence look like? Counts of entry-level vacancies over time, application volumes, employer hiring plans, graduate employment outcomes and comparisons between roles that use AI heavily and those that do not. Without such material, a single graduate’s search can illustrate a debate but cannot adjudicate it.

Advice without guarantees

Sheikh would still recommend computer science to someone who loves the technology and wants to code, but she would advise starting networking, communication and application skills earlier and being realistic about opportunities and the job market. She also urged employers to keep hiring entry-level talent to build the next generation of professionals, while acknowledging that AI’s impact remains uncertain. That appeal is a judgement about future capacity, not a finding from the data.

Her curriculum point is similarly suggestive rather than conclusive. She said she felt she had graduated a little too early because AI would probably become more integrated into learning, and that AI-related material she encountered from some professors was not part of the curriculum. That is one graduate’s recollection, not a curriculum survey, and no university response was available in the reporting. It raises a question about how quickly courses should adapt; it does not answer it.

The practical takeaways are therefore modest. Sheikh’s advice points towards early networking, communication and application practice as ways to improve a graduate’s chances, not as guarantees. Building websites, volunteering and fellowships formed part of her search, but one case cannot show that these activities caused the offer or that they are a general solution. Living at home gave Sheikh a financial cushion that not every graduate has.

The question that remains

The most useful lesson in Sheikh’s case is not that AI did or did not cost her a job, or that her eventual offer vindicates optimism about the market. It is that both the optimistic and pessimistic stories are being told with more confidence than the evidence supports. Her search documents difficulty, persistence and a good outcome. It leaves open whether entry-level pathways are narrowing, how curricula are keeping pace and what evidence would settle the argument.

Until better evidence arrives, her experience is best read as a case study in uncertainty. It shows why individual stories matter: they give a human shape to a labour-market argument. It also shows why they are not enough. The evidence does not show whether the first job in the AI era is harder to find, or whether the market is simply changing. One graduate’s successful search cannot tell us which, and neither can a column that treats her outcome as a lesson for everyone.

Sheikh chose to specialise rather than switch subjects, betting that deeper understanding would outlast the tools of the moment. That instinct travels well beyond a job search. Networks, protocols, how weaknesses are described and how authorised testing probes them are the parts of security that stay useful when a platform, a model or a vacancy list changes underneath you, because they let you reason about a system rather than simply operate today’s version of it. Anyone weighing a similar investment in durable knowledge can explore The Hack Academy’s online cybersecurity training programme, which lists cybersecurity fundamentals, networking and internet vulnerabilities among its subject areas, and sets self-paced learning and practical exercises alongside the theory.