Seth Walsh
Iconoclast
Contributor
- Joined
- Jan 12, 2020
- Posts
- 10,500
- Reputation
- 21,410
THE REAL REASON YOU CAN’T FIND WORK
The entry-level ladder is being removed one rung at a time.
AI did not create the crisis. It is accelerating a system that already stopped training outsiders.
The entry-level ladder is being removed one rung at a time.
AI did not create the crisis. It is accelerating a system that already stopped training outsiders.
You were told the problem was your CV, attitude, interview technique or lack of applications.
Those things matter. They are not the root cause.
The real problem is that the labour market increasingly demands evidence which a beginner can only acquire after somebody has already hired him.
The first rung used to be inefficient but available: junior administration, basic analysis, routine coding, local retail, apprenticeships, trainee roles and low-stakes office work. Employers paid beginners to perform simple tasks while learning the difficult ones.
That bargain is breaking.
- Software and AI compress routine tasks.
- Employers carry less spare labour and less training capacity.
- Online applications produce an ocean of interchangeable candidates.
- Credentials have become common enough to screen with, but weak enough not to trust.
- Housing and transport costs prevent workers from moving toward opportunity.
- Class and networks decide who can wait, relocate, intern and get vouched for.
You are not merely competing with other unemployed people. You are competing with experienced workers, internal candidates, referrals, contractors, software—and the option to leave the position unfilled.
I. THE NUMBERS SAY THE FIRST RUNG IS FAILING
The supplied ITV report covers Britain’s youth-employment crisis. The official ONS estimate for January–March 2026 is stark:
13.5%
of UK 16–24-year-olds were not in education, employment or training.
of UK 16–24-year-olds were not in education, employment or training.
That is more than one million young people.
The report also says:
- six in ten young people outside work or training have never had a job, versus four in ten in 2005;
- apprenticeships in England have fallen 35% since 2017;
- there are 1.6 million fewer low- and medium-skilled jobs than twenty years ago;
- one in three of the excluded young people has good GCSEs and one in six has a degree.
This is not a simple qualifications shortage. The country has educated more people while weakening the bridge that turns education into occupational experience.
A degree may raise lifetime opportunity. It does not force an employer to absorb the cost and uncertainty of converting a graduate into a useful worker.
II. CAPITALISM DOES NOT OWE YOU AN APPRENTICESHIP
Strip away the moral language. A firm is not a school, family or rehabilitation service. It is an organisation trying to produce more value than it consumes.
Hiring a beginner creates four costs:
- Wage cost: money paid before full productivity.
- Training cost: productive workers must stop producing to teach.
- Error cost: beginners create rework, supervision and reputational risk.
- Exit risk: once trained, the worker may leave for a better employer.
When labour was required for large volumes of routine work, firms tolerated those costs. The routine work paid for the training period.
When technology removes the routine layer, the trainee becomes a pure investment with an uncertain return.
Sometimes the substitute is a machine. Sometimes it is AI. Sometimes it is an experienced employee doing two roles. Sometimes it is an overseas contractor. Sometimes it is simply not doing the work.
This is why “there are vacancies” can be true while you remain unable to get hired. Aggregate demand for labour is not the same thing as demand for your current bundle of skill, location, proof and risk.
III. AI IS DESTROYING THE APPRENTICESHIP LAYER
The low-IQ interpretation is that AI will instantly replace every worker.
The better interpretation is that AI changes the task composition of jobs before it changes the number of jobs.
The ILO’s 2025 occupational-exposure index estimates that one in four workers globally is in an occupation with some exposure to generative AI. Only 3.3% of global employment falls into its highest exposure category. The ILO concludes that job transformation is more likely than complete replacement because most occupations still contain tasks requiring human input.
That sounds reassuring. For beginners, it may not be.
Imagine an analyst’s job contains:
- 40% gathering information;
- 30% cleaning and formatting it;
- 20% drafting routine commentary;
- 10% exercising judgment with a client.
The first three categories were how a junior earned his seat while slowly learning the fourth. If AI allows one senior analyst to compress that work, the final judgment remains human—but the junior position may vanish.
AI can transform a profession while deleting its training route.
This is the part most “AI will create new jobs” arguments miss. New jobs can emerge in the aggregate while the conversion process from novice to trusted operator becomes narrower, more selective and more class-dependent.
The winner is not “the person who uses ChatGPT.” Everyone can use ChatGPT.
The winner is the person who possesses domain knowledge, owns the problem, checks the output, carries liability and uses AI to multiply an already credible skill.
IV. THE INTERNET TURNED HIRING INTO A SPAM WAR
Online applications reduced the cost of applying toward zero.
That sounds democratic. It also produces noise.
Candidates respond to low response rates by sending more applications. Employers respond to more applications with harder filters, automated screening and preference for familiar signals. Candidates then optimise keywords and use AI to generate more applications. The signal-to-noise ratio falls again.
LOW RESPONSE → MASS APPLYING → MORE NOISE → HARDER FILTERS → LOWER RESPONSE
This explains the modern humiliation ritual:
- upload the CV;
- retype the CV;
- write a personalised statement;
- complete a personality test;
- record a one-way interview;
- receive an automated rejection—or silence.
The system is not carefully evaluating each person and declaring him worthless. Most candidates are barely being evaluated at all.
You are inside a congested distribution channel, not a meritocratic tournament.
Sending another 300 generic applications through the same channel is not persistence. It is repeating a low-yield trade.
V. SOCIAL CLASS DETERMINES WHO CAN SURVIVE THE QUEUE
Two graduates can have the same intelligence and qualification while entering completely different labour markets.
The upper-middle-class beginner often has:
- a quiet home and functioning laptop;
- parents who can cover six months without income;
- access to a car or money to relocate;
- familiarity with professional speech and dress;
- friends whose parents work in institutions;
- permission to accept low-paid prestige work;
- someone who can explain which credentials matter and which are scams.
The working-class beginner often needs the first job to pay immediately.
He cannot complete a three-month unpaid placement, move to London on speculation, wait through a five-stage interview or accept a low salary in exchange for future signalling value. His rational short-term choices can therefore damage his long-term trajectory.
This is not about rich people being genetically superior or poor people lacking ambition. It is about runway, information and trusted introductions.
Opportunity Insights analysed 21 billion Facebook friendships. It found that economic connectedness—the share of higher-SES friends among lower-SES people—was among the strongest predictors of upward mobility identified in the study. Its counterfactual estimate suggests that giving lower-SES children the connectedness of higher-SES children would raise their adult incomes by 20% on average.
A referral is not merely nepotism. It is privately transferred trust.
The open application asks a stranger to believe your claims. The referral makes someone else stake part of his reputation on them.
That is a different product.
VI. PATH DEPENDENCY: YOUR FIRST JOB CHANGES YOUR NEXT JOB
Careers are not clean ladders. They are branching paths.
FIRST ROLE → SKILL → REFERENCE → NETWORK → BETTER ROLE → STRONGER SIGNAL
NO FIRST ROLE → GAP → WEAKER SIGNAL → WORSE OPTIONS → LONGER GAP
NO FIRST ROLE → GAP → WEAKER SIGNAL → WORSE OPTIONS → LONGER GAP
The first outcome changes the probability of the next one. Economists call the long-term damage from early unemployment scarring. OECD work notes that early youth unemployment can damage earnings and employment opportunities even twenty years later.
Why?
- skills decay or never form;
- references remain absent;
- professional networks do not develop;
- confidence and routine deteriorate;
- employers treat the gap as information;
- the candidate lowers his target and enters a weaker track.
This is path dependency. The market does not evaluate your timeless potential. It observes the evidence produced by your previous path.
The cruelest sentence in hiring is “experience required.” The experience itself was allocated by an earlier hiring decision.
VII. THE BEHAVIOURAL TRAP IS REAL—BUT IT IS NOT THE ROOT CAUSE
The ITV report describes young people sleeping at 2, 3, 4 or 5 a.m. and living in their bedrooms while doomscrolling.
This behaviour is both a consequence and an amplifier.
Rejection removes structure. No structure destroys sleep. Bad sleep weakens attention, mood and interview performance. More rejection follows. The phone provides immediate reward with no risk of humiliation.
REJECTION → WITHDRAWAL → DYSREGULATION → LOWER CAPACITY → MORE REJECTION
Calling this laziness is useless. Pretending behaviour does not matter is equally useless.
The labour market may be structurally hostile. You still need to arrive rested, coherent, socially calibrated and capable of sustained work. Structural analysis should improve strategy—not become a sedative.
VIII. THE COPES THAT KEEP YOU UNEMPLOYED
- “I just need more applications.” Not if the channel is saturated and your evidence is unchanged.
- “I’ll get another generic degree.” Education without a defined hiring gate can become expensive delay.
- “AI skills.” Prompting without domain competence is not a profession.
- “Personal branding.” An audience cannot substitute for proof unless attention itself is the product.
- “Remote only.” Beginners voluntarily rejecting geographic and in-person opportunity are choosing the most competitive queue.
- “I’m above that role.” Prestige is less important than whether the role produces skill, references and proximity to better work.
- “Capitalism is broken, so nothing matters.” A system can be structurally unfair and still punish bad strategy.
- “I’ll wait until I feel confident.” Confidence usually arrives after evidence, not before it.
IX. HOW TO BREAK INTO THE MARKET THAT ACTUALLY EXISTS
01 — PICK A PROBLEM, NOT AN IDENTITY
Do not say “I want to work in business,” “tech,” “media” or “finance.” Name a recurring problem an employer pays to remove.
- recover overdue invoices;
- book qualified sales calls;
- maintain a compliance process;
- analyse a specific dataset;
- repair a specific machine;
- reduce customer response time;
- produce a defined creative asset;
- coordinate a regulated workflow.
The narrower the problem, the easier it is to build relevant proof.
02 — REPLACE CLAIMS WITH ARTIFACTS
A CV says you can. An artifact lets the employer inspect the claim.
Build three pieces of evidence:
- a completed project using realistic constraints;
- a short written explanation of the decisions and trade-offs;
- a reference, testimonial or measured outcome.
Use AI to accelerate research, drafts, testing and repetitive work. Then verify everything manually and explain it without AI. If you cannot defend the output, it is not proof of your competence.
03 — LEAVE THE OPEN-APPLICATION CASINO
Open applications should be one channel, not your entire strategy.
- Target firms before they advertise.
- Contact the person who owns the problem, not generic HR.
- Use local and in-person markets where the applicant pool is smaller.
- Approach smaller firms that cannot afford elaborate recruitment.
- Ask practitioners what gets a junior trusted in their field.
- Follow up with new evidence, not “just checking in.”
04 — ENTER AN INSTITUTION
Courses are useful when they provide a recognised gate, equipment, placement, licence, apprenticeship or repeated contact with employers.
Useful institutions include:
- trade and professional bodies;
- apprenticeship providers;
- competitive sports and civic organisations;
- serious volunteer operations;
- industry events with recurring attendance;
- local employers with internal promotion routes.
Attendance creates familiarity. Service creates trust. Responsibility creates references.
05 — BUILD REFERRALS BY BECOMING LOW-RISK
Do not ask strangers for jobs. Give people enough evidence that recommending you becomes safe.
- be specific;
- do small promises exactly on time;
- write clearly;
- make useful introductions;
- keep confidences;
- show improvement after criticism;
- ask for advice, apply it, then report the result.
The objective is not to “know important people.” It is to become somebody competent people are comfortable attaching their name to.
06 — BUY RUNWAY
Class disadvantages cannot be deleted by budgeting, but a lower burn rate widens your strategic options.
- cut consumer debt and status spending;
- live with family or share housing when the trade-off is rational;
- work temporary hours while protecting time for the target skill;
- widen geography;
- own reliable transport where it materially expands the market;
- avoid locking yourself into costs that require immediate bad employment.
Runway converts desperation into selectivity.
07 — PROTECT MOMENTUM AT ALL COSTS
If the ideal role is unavailable, choose work that produces at least one of the following:
SKILL · REFERENCE · LICENCE · PORTFOLIO · NETWORK · CASH · ROUTINE
A bridge job is not failure if it deliberately moves you toward the next gate. A dead-end job is one that consumes all capacity while producing none of those assets.
X. THE 30-DAY ESCAPE PROTOCOL
DAYS 1–3: DIAGNOSE
- Choose one role family and one geography.
- Collect 30 real job descriptions.
- List the five repeated problems, tools and proof requirements.
- Identify which gate is blocking you: skill, proof, network, location or signal.
DAYS 4–10: BUILD PROOF
- Complete one realistic artifact addressing the most common problem.
- Document the process in one page.
- Get it criticised by three practitioners.
- Revise it until you can explain every decision.
DAYS 11–20: CREATE PROXIMITY
- Contact two relevant practitioners per day with a specific question.
- Attend two in-person industry or institutional events.
- Offer one bounded piece of useful work.
- Ask which employers genuinely train beginners.
DAYS 21–30: CONVERT
- Apply only where your proof matches the problem.
- Send the artifact directly to the problem owner.
- Ask trusted contacts for introductions, not endorsements they cannot honestly give.
- Practise explaining your work without jargon.
- Track conversations, interviews and objections—not raw application count.
THE LAST WORD
You cannot individually reverse automation, rebuild apprenticeships, lower housing costs or make employers train again.
You can understand the actual game.
The system is selecting for people who arrive with proof, context, runway and transferred trust. That selection is profoundly shaped by capitalism and social class. AI raises the standard further by making routine beginner output cheap.
It is not all your fault. It is still your move.
Stop measuring effort by applications sent. Build scarce competence. Produce inspectable evidence. Enter institutions. Become useful to serious people. Protect routine. Lower your burn rate. Move before the gap becomes your identity.
THE LADDER IS NARROWER.
CLIMB DELIBERATELY.
CLIMB DELIBERATELY.
- ITV News — Why these young people are struggling to find work (28 May 2026)
- Office for National Statistics — Young people not in education, employment or training, May 2026
- International Labour Organization — Generative AI and Jobs: A Refined Global Index of Occupational Exposure (2025)
- Opportunity Insights / Nature — Social Capital I: Measurement and Associations with Economic Mobility (2022)
- OECD — Setting objectives for achieving better youth employment outcomes
Country- and year-specific figures are labelled. Exposure estimates describe tasks and occupations, not guaranteed individual job loss.
