Software engineering once offered a clear path upward. New graduates joined teams, handled routine tasks, learned from seniors and gradually took on complex work. Those days look different now. A veteran engineer who spent 14 years at Microsoft, Twitter and Stripe sees AI changing the equation in real time. “I think junior engineers could be affected the most,” he told Business Insider.
Manoj Aggarwal shared his perspective in a first-person account published Friday. He described how AI coding assistants have accelerated his own output dramatically. Prompts replace long stretches of manual coding. Projects that once demanded weeks now move at speeds he compares to 10 times faster. Yet the accountability stays with humans. Review of AI-generated code adds new layers of scrutiny. Burnout risks rise even as productivity climbs.
The pattern extends beyond one engineer’s desk. Entry-level postings in the United States have dropped 35 percent in the past 18 months, driven largely by artificial intelligence taking over foundational work. The World Economic Forum highlighted the trend in March, citing research from Revelio Labs. Routine tasks that once trained young professionals now flow directly to machines. What remains pushes upward. Managers absorb the load. Seniors stretch to cover gaps once filled by juniors.
But not every organization follows the same script. Some companies report that AI actually increases demand for entry-level hires. A May survey covered in The Wall Street Journal found nearly three times as many executives at AI-using firms planned to expand junior hiring in 2026 as those intending to cut back. The technology handles rote elements. New graduates step into roles with greater analytical responsibility from day one. Complexity replaces volume.
That shift carries consequences. Colleges watch their students lose traditional training grounds. “As AI automates routine work, students are losing the early-career roles where they developed skills like judgment and accountability,” Forbes reported in January. IMF Managing Director Kristalina Georgieva captured the scale at Davos. She called AI “like a tsunami hitting the labor market” and noted that entry-level positions often feel the first impact.
Numbers tell a stark story. Employment for software developers aged 22 to 25 has fallen 23 percent since ChatGPT launched in late 2022. The 26-to-30 group declined 5 percent. Workers aged 41 to 49 saw an 18 percent gain over the same period. A 41-percentage-point spread separates the youngest cohort from experienced peers. CNBC laid out the data last September, drawing from labor analysis that points to AI exposure as a key factor.
Venture capital firm SignalFire tracked hiring at major public tech companies and later-stage startups. New role starts for people with less than one year of experience dropped 50 percent between 2019 and 2024. The decline held steady across functions from engineering to marketing. Asher Bantock, head of research at SignalFire, described the figure as an accurate representation of the shift.
Aggarwal stays ahead by treating AI literacy as non-negotiable. He dedicates a couple of hours each week outside office demands. A toddler at home sets natural boundaries. Laptop time arrives only after bedtime or on weekends when the child sleeps. The schedule preserves balance while allowing exploration. “I think it’s important to spend a couple of hours a week outside work experimenting with AI to stay up to date with the industry,” he explained.
His monthly spend stays modest. Fifty to sixty dollars covers Claude Code usage. He experiments with Microsoft Copilot, Lovable and Anthropic’s latest models. One personal project produced an open-source tax and financial assistant. Users connect models such as Gemini or ChatGPT, upload documents and receive analysis of spending patterns, subscriptions and portfolio risk. Aggarwal ran his own 2025 taxes through the locally hosted tool. It confirmed his CPA’s work without exposing sensitive data online.
His message lands clearly. Domain knowledge and system-level thinking remain human strengths. AI lacks context about why architectures exist or how business goals shape technical decisions. “I’d suggest leaning into your experience — both your engineering skills and the institutional knowledge you’ve built if you’ve been at a company for a while,” Aggarwal advised. Human oversight stays essential because models still err.
Discussions on X reflect the tension. One post from mid-July cited the same generational employment gap and concluded that AI did not eliminate junior roles so much as eliminate tolerance for paying someone to learn on the job. “The industry is eating its own pipeline and calling it efficiency,” the user wrote. Another thread listed junior software engineers alongside customer service and QA testers as positions already handled by AI in many organizations.
Recent coverage adds nuance. A July 25 article on LatestLY echoed Aggarwal’s warnings under the headline “Junior Engineering Jobs at Risk.” It amplified concerns that cost-cutting favors AI investment over entry-level headcount. Yet conversations among developers on platforms like Hacker News push back. Some argue offshoring and preference for senior hires explain more of the drop than AI alone.
The debate matters for the profession’s future. Fewer juniors today means thinner benches tomorrow. Companies that flatten structures to cut costs may discover they lack the next generation of leaders who once rose through hands-on mentorship. Middle managers already report feeling overextended as AI-routine work migrates upward. Disengagement follows.
Aggarwal offers practical counsel. Engineers should test different tools. They should measure their own productivity gains against peers who ignore the shift. Those who adapt will likely command greater responsibility even if headcount shrinks. “Regardless of how you feel about AI, if you’re compared with another engineer who’s using it, there’s no doubt that the other engineer’s productivity will be higher,” he said.
Tech leaders face hard choices. They can automate entry points and risk starving their talent pipeline. Or they can redesign junior roles around higher-order skills that AI complements rather than replaces. The data suggests many have already chosen the first path. Entry-level opportunities evaporated. Young developers scramble. Seniors gain leverage.
Still, the picture isn’t uniform. Organizations that view AI as a multiplier rather than a substitute continue to hire graduates. They expect those new employees to analyze, decide and innovate from the start. The learning curve steepens. The safety net of menial tasks disappears. Success belongs to those who arrive prepared to operate at a senior level of thinking despite limited experience.
Universities confront parallel pressure. Career services offices must reconsider how they prepare students when traditional first jobs no longer exist in the same form. Some programs integrate AI fluency into curricula. Others emphasize judgment, communication and business acumen that machines cannot replicate. The bargain between education and employment frays.
Aggarwal built his side project to solve a concrete problem. He wanted accurate tax preparation without cloud exposure. The result demonstrates what remains possible when experienced engineers direct powerful models toward specific goals. His story offers both warning and roadmap. Junior roles may contract. Demand for thoughtful, accountable engineers who master AI as a collaborator will persist.
The industry stands at a fork. One direction leads to hollowed-out career ladders where only proven experts thrive. The other builds teams that pair human insight with machine scale at every level. Which path prevails will shape software development for decades. Early evidence points to contraction at the bottom and intensification at the top. Engineers who recognize the change and act on it position themselves on the right side of that divide.
WebProNews is an iEntry Publication