The ladder into a professional career used to have a reliable bottom rung. You graduated, you took a junior role that did not ask much, and you learned the work on someone else's payroll. That rung is not what it was. For anyone standing below it, or watching a son or daughter stand below it, this is what happened and what the people who are still climbing are doing differently.

The measurement

The clearest number comes from SignalFire, which tracked hiring at the fifteen largest technology companies. New-graduate hiring at those firms is down more than 50% since 2019 (SignalFire State of Tech Talent Report, 2025). The composition shifted just as sharply: before the pandemic, recent grads made up about 15% of hires at big tech; by 2024 they were roughly 7%, half the earlier share (SignalFire State of Tech Talent Report, 2025).

Two forces are pressing at once. The first is ordinary economic caution; uncertain employers hesitate to invest in people who need training. The second is more specific to this moment: AI is absorbing exactly the kind of routine, learn-on-the-job tasks that used to be handed to interns and first-year hires (Fortune, Aug 15, 2025). The rung did not just get smaller. Part of what made it a rung, the pile of simple work a beginner could cut their teeth on, is being done by a tool.

SignalFire named the trap that results the "experience paradox": companies post a junior req and then fill it with someone who has years of experience, skipping the actual beginner entirely (SignalFire State of Tech Talent Report, 2025). The listing says entry-level. The hire is not. For a genuine new entrant, that is the wall behind many quiet rejections.

Why this is not a life sentence

It would be easy to read those numbers as a closed door, and for a young person especially, that reading is corrosive. So it is worth being precise about what the data does and does not say. It says the traditional front door, the campus-recruiting pipeline into a big employer, has narrowed. It does not say early-career people have stopped getting hired. It says the path changed shape.

The same PwC research that documents disruption also documents where the value is moving. AI-exposed entry-level roles are now seven times more likely to ask for skills once considered senior, such as judgment and leadership (PwC 2026 Global AI Jobs Barometer, Jun 2026). That is a harder bar, but it is also a hint. The beginner who can show judgment, not just willingness, is being asked for, and that is a thing a person can deliberately build and demonstrate before anyone hires them.

The workarounds becoming a playbook

The people climbing anyway are not waiting for the old rung to come back. They are assembling evidence of capability outside the traditional pipeline, and a few patterns recur often enough to call a playbook.

They build in public. When the front door narrows, demonstrated work substitutes for the credential the front door used to confer. SignalFire's own read of the new path points to bootcamps, open-source contributions, freelancing, and self-directed projects as the routes that are working (SignalFire State of Tech Talent Report, 2025). A portfolio of real, visible output does what an entry-level job used to do: it proves you can do the work before a company has to bet on you.

They enter through people, not portals. The referral advantage is not a nicety here; it is the mechanism that gets a real beginner past the experience paradox. Referred candidates are seven times more likely to be hired than those who apply through job boards (Pinpoint, 2023). A referral is, in effect, a person vouching that you can do the work despite a thin resume, which is exactly the gap a new entrant needs bridged. For someone without a network yet, the move is to build one deliberately through communities, open-source projects, and the connections a portfolio attracts.

They target the edges, not the center. The fifteen largest tech firms are where the contraction is sharpest. Smaller companies, non-tech employers hiring for AI-adjacent roles, apprenticeship-style programs, and functions like marketing, operations, and finance that are now hiring for AI fluency without a coding requirement (PwC 2026 Global AI Jobs Barometer, Jun 2026) are less saturated and more willing to take a promising beginner. The center of the market is the hardest place to enter; the edges are where the rung still exists.

For the person on the ground

If you are early in your career, or supporting someone who is, the honest frame is neither denial nor despair. The traditional entry-level job is scarcer and the numbers are not spin. But scarcer is not gone, and the path around the missing rung is now well enough traveled to follow.

The practical version is short. Produce visible work that proves capability, because demonstrated skill is now the credential. Enter through relationships, because a referral is what carries a thin resume past the experience paradox. And aim past the fifteen biggest names, where the door has narrowed most, toward the many employers where a capable beginner is still exactly what is wanted. The rung broke. People are still climbing. They are just building their own footholds on the way up.