Entry-level positions dropped 60% since 2022. Here's the brutal reality—and the unexpected pathways forward—for aspiring developers in the AI era.
As a recent Software Engineering graduate navigating London's tech market, I've witnessed firsthand what the data confirms: securing that first developer role is considerably more challenging than it was just three years ago. The figures are stark and warrant an honest discussion.
Entry-level software development roles have dropped by sixty per cent between 2022 and 2024. Tech internship postings fell by thirty per cent during the same period, whilst applications increased by seven per cent. Computer science graduates now face unemployment rates exceeding seven per cent—well above the national average and significantly higher than fields such as nursing or engineering.
Whilst artificial intelligence coding assistants grab headlines as a major factor, they're not the only culprit. The market correction following the pandemic hiring boom, economic uncertainty, rising interest rates affecting tech funding, and companies prioritising profitability over growth have all contributed to this perfect storm. AI simply became the convenient scapegoat for broader economic shifts.
This isn't another pessimistic piece about AI replacing humans. The reality is far more complex and, surprisingly, contains genuine reasons for optimism. The industry isn't eliminating junior positions—it's fundamentally reimagining what they represent and who succeeds in them.
Between 2020 and 2022, tech companies went on unprecedented hiring sprees, fuelled by cheap money, rapid digitalisation, and pandemic-driven demand. Many organisations over-hired dramatically, leading to the mass redundancies we've seen throughout 2023 and 2024.
This correction hit junior positions hardest. When companies needed to trim headcount, they protected experienced engineers who could immediately contribute whilst cutting roles designated for training and development. The pendulum swung from "hire everyone" to "freeze all non-essential positions" almost overnight.
Rising interest rates transformed the tech funding landscape. Startups that previously raised capital easily now face intense scrutiny over profitability and burn rates. Venture capital funding for tech companies dropped substantially, forcing organisations to operate more conservatively.
Junior developers, requiring months of training before becoming productive, became luxuries rather than necessities in this environment. Companies facing survival pressures naturally prioritise immediate value over long-term talent development.
Yes, AI plays a significant role—but perhaps not quite as dramatically as headlines suggest. Activities that once consumed a new developer's first year—generating boilerplate code, writing basic functions, debugging simple errors, creating documentation—can now be automated.
Major tech companies have responded accordingly. Salesforce announced it would hire no new engineers in 2025, citing AI productivity gains. Google and Meta reduced new graduate hiring by roughly fifty per cent compared to 2021 levels. When GitHub Copilot costs £8 monthly versus a junior developer's £35,000-50,000 annual salary, the economic equation appears straightforward.
Coding bootcamps proliferated during the pandemic, promising rapid career transitions. Simultaneously, universities expanded computer science programmes in response to perceived demand. The result? A massive influx of entry-level candidates competing for shrinking position numbers.
This oversupply gives employers leverage to demand increasingly unrealistic requirements. Positions labelled "entry-level" now routinely require two to five years of experience, multiple technology stacks, and portfolio projects demonstrating production-level quality.
The skills considered "foundational" for junior developers have evolved rapidly. It's no longer sufficient to know one language and framework. Employers expect familiarity with cloud platforms, containerisation, CI/CD pipelines, version control, testing frameworks, and increasingly, AI tool proficiency—all before the first job.
This creates a chicken-and-egg problem: you need experience to get hired, but you need to get hired to gain experience. The traditional pathway of learning on the job has narrowed considerably.
Survey data reveals a troubling disconnect. Seventy per cent of hiring managers believe AI can perform intern-level work, with fifty-seven per cent trusting AI outputs more than recent graduate contributions. Thirty-seven per cent openly state they'd rather deploy AI than hire recent graduates.
This perception creates a harmful cycle. As companies hire fewer juniors, fewer professionals gain the foundational experience needed to become mid-level and senior engineers. Industry veterans warn that organisations failing to develop talent pipelines will face severe shortages within five to ten years—but short-term cost pressures often override long-term workforce planning.
Not everyone accepts the premise that AI makes junior developers redundant. AWS CEO Matt Garman called this notion "one of the dumbest things I've ever heard," questioning how companies expect to cultivate senior talent without entry-level hiring. His message to the industry: continue recruiting and training graduates at historical levels.
Google CEO Sundar Pichai frames AI differently—as a productivity enhancer rather than replacement technology. By eliminating tedious tasks, AI allows engineers to focus on higher-impact work. His observation: increased productivity means Google can accomplish more with existing teams and potentially expand hiring as new projects become viable.
Former GitHub CEO Thomas Dohmke emphasises that junior developers bring fresh perspectives, diverse backgrounds, and innovative thinking that AI cannot replicate. GitHub continues operating robust internship programmes, demonstrating commitment to developing early-career talent despite AI capabilities.
Here's where the narrative shifts dramatically. My generation enters the workforce already fluent in AI-assisted development. Rather than spending weeks memorising syntax, we leverage tools like GitHub Copilot or ChatGPT to contribute meaningfully from day one.
Research indicates developers using AI assistants complete tasks up to fifty-six per cent faster, with junior developers experiencing the most significant productivity gains. We're not competing against the technology—we're the first generation to truly collaborate with it naturally.
This represents a fundamental advantage. Whilst senior developers must retrofit workflows built over decades without AI assistance, recent graduates naturally integrate these tools into our problem-solving approaches. We mightn't hand-code complex algorithms from scratch, but we excel at prompting AI to scaffold solutions whilst focusing energy on higher-level logic, architecture, and integration challenges.
The industry is creating entirely new positions that didn't exist three years ago. Companies now seek AI-augmented developers, prompt engineers, AI systems reviewers, automation specialists, and AI integration consultants. These roles typically combine development fundamentals with AI literacy—precisely the skillset recent graduates can cultivate.
AI specialist positions have grown at a seventy-four per cent annual rate over the past four years. Job postings requiring generative AI skills in non-tech roles increased ninefold from 2022 to 2024. Developers who position themselves at the intersection of traditional software engineering and AI orchestration find expanding opportunities rather than contracting markets.
These aren't just tech companies. Financial services, healthcare, retail, logistics, and government sectors all seek developers who understand both traditional programming and AI integration. Your ability to bridge these worlds becomes increasingly valuable.
Understanding the landscape is crucial, but actionable strategy matters more. Based on research, industry insights, and personal job-hunting experience, here are the capabilities that distinguish successful junior developers in the AI era.
Don't just use AI coding assistants—become expert at leveraging them. This means understanding how to craft precise prompts, recognize when AI suggestions are problematic, and know when to bypass automation entirely.
Practical approach: Use GitHub Copilot, ChatGPT, Claude, or similar tools extensively during personal projects. Document situations where AI helped versus hindered. Develop intuition about these tools' strengths and limitations.
Key insight: The goal isn't to compete with AI at writing code—it's to become irreplaceable at directing AI effectively and validating its outputs.
AI-generated code compiles and passes superficial tests while containing subtle bugs that manifest in production. The ability to spot edge cases, identify security vulnerabilities, and recognize logical flaws in AI outputs is increasingly valuable.
One engineering leader shared a revealing anecdote: during interviews, they provide candidates with clean-looking AI-generated code containing a subtle concurrency issue. Most applicants proceed without question. The few who pause to analyze and flag the problem demonstrate exactly the mindset modern teams value.
How to build this skill: Never blindly trust AI-generated code. Always review it line by line, considering edge cases, performance implications, and potential failure modes. Treat AI as a helpful but fallible teammate requiring oversight.
Writing individual functions is commoditized. Understanding how code connects to broader architecture, user experience, and operational impact remains deeply human work. Companies prioritize candidates who think beyond isolated tasks to consider system-wide implications.
This involves asking questions like: How does this feature affect database performance at scale? What security considerations arise from this API design? How will this decision impact future maintenance?
Development strategy: When building projects, force yourself to make architectural decisions. Document your reasoning. Consider alternatives and tradeoffs. This demonstrates maturity beyond years of experience.
Technical skills alone no longer suffice. The ability to communicate clearly with both technical and non-technical stakeholders, understand business requirements, and translate between domains increasingly differentiates candidates.
This includes writing effective prompts for AI (which requires clarity and precision), explaining technical decisions to product managers, and collaborating across functional teams. Emotional intelligence and interpersonal skills matter more than ever.
Practical application: Practice explaining technical concepts to non-technical friends or family. Write clear documentation for your projects. Engage in open-source collaborations requiring coordination with distributed teams.
Don't hide AI usage in portfolio projects—showcase it strategically. Demonstrate that you can leverage AI tools while maintaining code quality and understanding. Document where AI assisted and how you validated its contributions.
The strongest portfolios display: working applications deployed to live URLs, clean, well-documented code with clear architectural decisions, evidence of problem-solving beyond what AI alone could accomplish, and thoughtful integration of AI assistance rather than blind dependence.
Personal example: My Skill Maps platform combines React, Node.js, and PostgreSQL with JWT authentication and progress tracking. I used AI assistance for boilerplate generation and troubleshooting, but the architecture, security considerations, and feature design required human judgment and domain knowledge.
AI can generate algorithms, but understanding data structures, complexity analysis, design patterns, and software engineering principles remains essential. These fundamentals enable you to guide AI effectively and recognize when its suggestions are suboptimal.
Microsoft CEO Satya Nadella emphasizes that computational thinking ability matters tremendously, even as AI handles code generation. The developers who thrive understand the "why" behind the code, not just the "how."
Learning approach: Use AI as a learning accelerator, not a replacement for understanding. When AI generates a solution, study it thoroughly. Research alternative approaches. Understand tradeoffs and edge cases. Active learning combined with AI assistance creates powerful skill development.
Traditional routes into software engineering are narrowing, but alternative pathways are expanding. Here's what I've observed whilst navigating the job market, supplemented by strategies that have worked for others in similar situations.
Positions in IT support, quality assurance, technical writing, DevOps, or systems administration provide exposure to development environments whilst building practical experience. These roles often have lower barriers to entry and can serve as excellent launching pads.
My current position as an IT Service Desk Analyst has proven surprisingly valuable. I've gained understanding of enterprise systems, learnt how real organisations use technology, developed troubleshooting skills, and built professional networks within the industry. This experience makes me a stronger developer candidate because I understand the full ecosystem, not just code in isolation.
Action steps: Look for IT support analyst, QA tester, technical support engineer, DevOps assistant, or systems administrator roles. Many companies promote internally from these positions into development.
Whilst major tech firms reduce junior hiring, smaller organisations (10-100 employees) often maintain simpler processes and more immediate needs. These companies frequently offer hands-on learning opportunities that larger corporations no longer provide.
These environments typically involve wearing multiple hats, rapid skill development, and direct exposure to business decisions. The trade-off—lower initial compensation and less structure—can be worthwhile for career launch.
Where to find them: AngelList, Otta, local tech meetups, university career services, and LinkedIn. Don't overlook companies outside the traditional "tech" sector—many retailers, financial services, and healthcare organisations need developers.
AI tools have paradoxically made it easier for beginners to deliver client work. Platforms like Upwork, Fiverr, PeoplePerHour, and Freelancer allow developers to build portfolios, gain practical experience, and earn income whilst job searching.
Start with small projects: WordPress customisations, simple web applications, automation scripts, or data analysis tasks. Build reputation gradually. Success requires strong client communication and ability to scope projects realistically.
Pro tip: Treat every freelance project as a learning opportunity and portfolio piece. Document your process, showcase results, and request testimonials.
Contributing to established open-source projects provides visible evidence of collaboration skills, code quality, and ability to work within existing architectures. Many companies value open-source contributions as highly as formal employment history.
Start with documentation improvements or bug fixes before tackling features. Projects labelled "good first issue" or "beginner-friendly" welcome newcomers. Regular contribution demonstrates consistency and commitment.
Getting started: Explore GitHub topics in your preferred language, check out First Timers Only, and join project Discord/Slack communities. Code for Good and Hacktoberfest offer structured entry points.
UK companies increasingly offer software development apprenticeships—structured programmes combining work with formal training. These provide salary whilst learning and often lead to permanent positions.
Graduate schemes at larger organisations, whilst competitive, offer comprehensive training programmes. Companies like BT, Lloyds Banking Group, HMRC, and consulting firms run substantial graduate intakes annually.
Resources: Gov.uk Find an Apprenticeship, RateMyApprenticeship, TargetJobs, and company career pages. Applications often open 6-12 months before start dates.
Teaching programming—whether through bootcamps, online platforms, or private tutoring—reinforces your own knowledge whilst building credibility. Platforms like Codementor, Tutor Hunt, or local coding bootcamps seek instructors.
Creating technical content (blog posts, YouTube tutorials, or courses on Udemy) demonstrates expertise and can generate passive income. It also improves communication skills and builds your personal brand.
Bonus: Teaching forces you to understand concepts deeply enough to explain them simply—a skill that serves you throughout your career.
Rather than competing in oversaturated generalist markets, consider specialising in areas where demand exceeds supply:
Strategy: Choose a niche aligned with your interests, then build targeted expertise through projects, certifications, and networking within that community.
Many positions never reach job boards. Attend local tech meetups, hackathons, and conferences. Join communities like Silicon Drinkabout (London), CodeNewbie, or framework-specific groups (React London, Python UK, etc.).
LinkedIn networking isn't just about connection requests—engage meaningfully with content, share your learning journey, and contribute to discussions. Genuine relationships often lead to referrals and opportunities.
Practical approach: Attend one event monthly, follow up with 2-3 interesting people you meet, and offer value before asking for favours. Quality connections matter more than quantity.
If you're already employed in a non-technical role, investigate internal transfer opportunities. Many large organisations run retraining programmes to move employees into technical positions—particularly in finance, government, and healthcare sectors.
Companies often prefer internal candidates because they already understand the business context and culture. The path from business analyst, project manager, or operations role into development is more common than you might think.
Create and launch actual products—even simple ones. A deployed application with real users demonstrates capabilities far better than tutorial projects. Consider:
My Skill Maps platform exemplifies this approach—it's a working product that demonstrates full-stack capabilities, architectural decisions, and ability to ship complete solutions.
Key point: Getting users (even just a few) and iterating based on feedback shows product thinking—a skill employers increasingly value.
Whilst direct applications often disappear into ATS black holes, recruitment agencies can provide advocacy and access. Agencies like TEKsystems, Harvey Nash, Hays Technology, and Experis actively place junior developers.
Newer platforms like Otta, Cord, and Hackajob use better matching algorithms than traditional job boards. They often highlight growing companies more willing to invest in junior talent.
From experience: Build relationships with 2-3 specialist tech recruiters. They can provide market insights, CV feedback, and sometimes access to unadvertised positions.
Don't limit your search to London or the UK. Many companies hiring remotely care more about skills than location. Platforms like Remote.co, We Work Remotely, and Arc.dev connect developers with global opportunities.
Consider markets where demand outstrips supply—some European countries, US companies open to UK timezones, or emerging tech hubs. Your multilingual capabilities (Romanian, English, Russian, German) are genuine competitive advantages here.
Despite current difficulties, several factors suggest the junior developer market may stabilize or even improve in coming years.
Companies cannot sustain themselves without developing junior talent. Today's entry-level engineers become tomorrow's technical leaders. Organizations that aggressively cut junior hiring face talent shortages within five to ten years—a timeline many executives are beginning to recognize.
AWS CEO Matt Garman's pointed question resonates: "How's that going to work when ten years in the future you have no one that has learned anything?" Forward-thinking companies understand this dynamic and maintain investment in early-career development despite short-term cost pressures.
As companies deploy AI more extensively, they're discovering its limitations. Generated code requires careful review, maintenance costs remain high, and complex projects still demand human expertise. The initial euphoria about AI replacing developers is giving way to more realistic assessment of its role as augmentation rather than replacement.
History suggests technological disruption creates more opportunities than it eliminates. The World Economic Forum projects that AI-related technologies will create approximately eleven million jobs while displacing nine million—a net gain of two million positions.
Many of these emerging roles will require development skills combined with new specializations. The transition period is challenging, but long-term prospects remain strong for adaptable professionals.
Right, let's be real for a moment. As someone actively navigating this market with a First Class Honours degree and an award-nominated final year project, I won't pretend it's easy. Getting that first developer role is genuinely challenging right now—perhaps the most difficult it's been in a decade.
I've applied to what feels like hundreds of positions. I've tailored applications for specific companies until my eyes glazed over. I've worked with multiple recruitment agencies, each with slightly different advice. I've continuously enhanced my skills, built projects, and expanded my portfolio. And yes, I've faced plenty of rejections and silence.
But here's what I've learnt through this process: the challenge isn't insurmountable, and every setback teaches something valuable. Each application rejection reveals what the market actually values versus what I assumed it wanted. Every conversation with recruiters uncovers new angles to explore. Every hour spent building Skill Maps strengthens not just my portfolio but my actual capabilities.
The developers succeeding aren't necessarily the most talented coders—they're the most adaptable problem-solvers. They leverage AI effectively whilst demonstrating irreplaceable human judgement. They position themselves at intersections of multiple skill sets rather than competing in oversaturated spaces. They're persistent without being desperate, confident without being arrogant.
The market has changed, absolutely. But opportunity hasn't vanished—it's just wearing different clothes than it did three years ago. And that's actually fine, because we can change our wardrobes too.
My strategy? Keep building, keep learning, keep adapting. Use AI tools extensively whilst ensuring I understand fundamentals deeply. Network genuinely rather than transactionally—people can smell desperation a mile off. Take whatever tech role offers learning opportunities, even if it's not the perfect development position. Your first job doesn't define your career; it's just the entry point.
To fellow junior developers in the trenches: I see you. It's tough out there. But we're not competing against AI—we're learning to dance with it. And that dance? That's the skill that'll carry us through the next decade of our careers.
The data paints an uncomfortable picture: junior developer hiring has contracted significantly. AI has automated many traditional entry-level responsibilities. The post-pandemic market correction hit hardest. Economic headwinds persist. The path into software engineering has narrowed compared to the boom years of 2020-2021.
But viewing this solely through a crisis lens misses crucial context. Every technological revolution throughout history has disrupted existing career pathways whilst simultaneously creating new ones. The difference now is the pace of change—it's happening faster than previous transitions, which makes it feel more threatening.
For aspiring developers, the message is clear but not simple: adapt thoughtfully or struggle unnecessarily. This doesn't mean abandoning fundamentals or blindly chasing every new trend. It means understanding that the skills valued today differ from those valued five years ago, and they'll differ again five years hence.
The industry absolutely needs junior developers. Not necessarily in identical roles or with the same responsibilities as before, but the fundamental requirement for developing talent pipelines hasn't disappeared. Companies ignoring this reality will face consequences—creating opportunities for those positioned to meet future needs.
What's changing is the definition of "junior." It's no longer sufficient to know syntax and basic algorithms. The new baseline includes AI tool proficiency, systems thinking, communication skills, and ability to learn continuously. That's a higher bar, yes, but it's also a more interesting and sustainable foundation for a career.
Is it harder than before? Undeniably. Is it impossible? Absolutely not. Is it unfair that the goalposts keep moving? Perhaps—but fairness isn't really the point. Adaptability is.
The developers who thrive won't be those with perfect grades or those who resist AI most stubbornly. They'll be the ones who view AI not as threat or saviour, but as the most powerful tool in their arsenal—one that elevates their capabilities rather than diminishing their value.
The future belongs to AI-native professionals who can orchestrate intelligent systems whilst providing the creativity, judgement, and strategic thinking that machines cannot replicate. That future is being built right now, by junior developers willing to embrace rather than resist the transformation. And honestly? That's quite exciting, even if it's also terrifying.
I'm actively job searching in London and documenting insights along the way. Let's connect and share strategies for breaking into development in the AI era.
What's your experience as a junior developer in 2024-2025?
Share your story on LinkedIn or GitHub—your insights help others navigate this transition.