The Graduate Dilemma: The squeeze on entry-level jobs

The current squeeze on entry-level jobs may signal concern for tomorrow's students.

The entry-level white-collar job was never glamorous. It was, however, essential. It turned graduates into professionals by forcing them to practise judgement on low-stakes work while someone senior checked the answers. That apprenticeship model is now under pressure because generative AI is unusually good at the very tasks that once justified junior headcount: drafting, summarising, reconciling, searching, formatting and producing the first pass.

This is not chiefly a story about degrees becoming worthless. It is about the first rung of professional development being hollowed out faster than firms and education systems can rebuild it. The risk is not that AI replaces all professionals. The risk is that it removes the training ground that produces them.

The solution is a redesign of entry roles, a retooling of assessment, and an education model that behaves less like a credential factory and more like an incubation hub that remains in a student’s life long after graduation.

The bargain that is breaking

For decades, white-collar careers offered a quiet bargain. Graduates accepted a period of low autonomy and repetitive work in exchange for a credible path to competence and status. Employers tolerated the inefficiency of training because the pipeline produced future managers, partners and leaders. The early tasks were dull by design. Beginners should not be trusted with decisions that can bankrupt firms or harm clients.

Generative AI has attacked that bargain at its softest point.

Research on task exposure shows that a substantial share of knowledge work can be automated or accelerated this decade. Unlike previous waves of automation, this one reaches deep into language, analysis and routine information processing. That happens to be what many juniors are paid to do.

The economics are stark. A careful graduate may produce a decent first draft. A model produces ten drafts and invites you to choose. The spreadsheet no longer needs a roving interpreter on standby; it has a chat window. Under cost pressure, the temptation is obvious. Trim the layer whose output resembles a machine.

Yet what looks efficient in a quarterly report can be destructive over a decade.

The real risk is hollowing, not replacement

Students feel the pressure viscerally. Why spend years qualifying as a lawyer, accountant or doctor if software will do the job better by the time you graduate? The anxiety is rational. It is also slightly misdirected.

Professions are not merely about producing answers. They are about carrying responsibility for those answers, explaining them, defending them and bearing liability when they are wrong. AI may improve the first pass. It does not automatically assume institutional trust.

The more immediate threat is that AI removes the practice field.

In many organisations, juniors learned judgement by doing controllable work. Draft the note. Build the model. Summarise the cases. Reconcile the numbers. Seniors corrected the output and, over time, novices internalised what mattered. If the first draft is always machine-made and firms respond by shrinking junior cohorts rather than redesigning them, a training debt accumulates. Capability quietly erodes.

Some employers are already adjusting hiring processes. Skills-based screening is more common. Experiential learning and internships carry more weight. AI tools are used in selection as well as in production. But surface changes in job adverts are easier than structural changes in training and progression. Removing degree requirements does not automatically widen access or improve capability unless assessment and management practices change too.

If firms simply delete the first rung, they will eventually discover they have saved money and weakened their own succession pipeline.

Inventing the new first rung

The old entry job was to produce the first pass. The new entry job is to own the workflow around the first pass.

That shift creates opportunity rather than extinction.

Graduates will increasingly find roles in AI supervision and model validation, where the objective is not to generate output but to test it for reliability, bias, security and compliance. Data ethics and model governance will matter as organisations formalise accountability around automated systems. Human-AI interface work will expand, shaping prompts, guardrails and escalation paths so that tools fit real organisational behaviour. Digital operations and workflow orchestration will grow as firms rewire processes around automation. Cybersecurity will expand because more automation means more attack surface.

These are not consolation roles. They sit close to decision-making, risk and accountability. They reward graduates who can reason clearly, communicate precisely and understand systems rather than simply produce documents.

For employers, this requires discipline. Graduate programmes should be redesigned around verification, exception handling and stakeholder judgement, not endless drafting. AI subscriptions are cheap. Institutional capability is not. Training only looks wasteful if you assume your organisation will not exist in ten years.

Education as an incubation hub

Education’s challenge is not that it teaches handwriting instead of prompting. It is that too much of it still rewards compliance over creativity and recall over reasoning, even as machines excel at recall.

The response cannot be cosmetic. It must be structural.

Schools and universities can act as incubation hubs in three ways.

First, embed real-world collaborative projects into the core curriculum, with employers co-designing problems and co-evaluating output. Students should graduate with evidence of having navigated ambiguity, trade-offs and accountability.

Second, scale structured micro-internships and short, supervised work sprints. These give students credible proof of capability and give employers low-risk pathways to assess talent before committing to full-time hires.

Third, extend the relationship beyond graduation. The degree should not be a one-off transaction. It should function as a lifelong learning subscription: refresh modules, alumni project studios, pivot programmes and employer-linked certifications that keep skills current in a labour market that evolves quickly.

Institutions that remain static will face declining trust. Those that reposition as capability partners rather than credential vendors will become more valuable.

A new social bargain

The graduate predicament feels unfair because it is shaped by forces outside any individual’s control. Technology compounds rapidly. Institutions adapt slowly. Young people must plan long before the consequences are clear.

But the answer cannot be nostalgia.

Employers must keep the ladder intact by redesigning entry roles rather than deleting them, and by investing in meaningful assessment rather than convenient proxies. Educators must behave like capability builders, not exam administrators. Parents and students must shift their focus from safe job titles to compounding capabilities.

The question is no longer which profession is safe. The question is which skills deepen over time and retain value when machines handle the routine.

In an AI-rich world, information is abundant. Judgement is scarce. If society allows the first rung of professional development to disappear, it will not simply disrupt careers. It will erode the process by which expertise itself is formed.

The ladder need not break. But it will not repair itself.

By Louay Aldoory

Louay Aldoory is the Founder of 1648 Capital. 1648 Capital is a corporate advisory and private markets platform partnering with founders, shareholders, and investors on complex growth, restructuring, and capital structuring initiatives. We combine strategic insight with execution discipline, supporting businesses from transformation through to institutional capital alignment.

The strategies presented are thematic and do not constitute investment advice (or advice of any kind). No assurance can be given that the objectives of the investment above strategies will be achieved; the strategies involve risk (including, without limitation, illiquidity risk) and may incur a loss on some or all capital deployed. The opinions expressed, or indeed the information or assumptions that underpin them, may contain errors, mistakes, or omissions; no assurance or warranty can be made as to the accuracy or completeness of this information, and readers should not place any reliance on this content to execute investment decisions or for any other purpose. Readers accept full responsibility for using this content and are kindly requested to consult with their professional advisor before making any investment decision related to the same.

CHANGE BEGINS WITH A CONVERSATION