The Market Is About To Tilt Back Toward Juniors
The last few months of hiring interns reshaped how I read engineering talent. A controversial take: over the next two to three years, the most asymmetric hires are not going to be the coveted senior engineers. They are going to be the juniors who learned to build with AI from day one, and the companies still filtering on years of experience and pedigree are going to miss the shift.
I have spent the last several months hiring interns. The process opened my eyes to a few things I think are about to become a much bigger story across engineering hiring. This is a controversial take, so I will say it plainly. Over the next two to three years, the market is going to shift from the extremely senior, extremely coveted engineer to the junior who has already rewired how they build.
What actually changed
Fluency with AI tooling among juniors right now is unlike anything I have seen before. The kids I have been interviewing are moving at a pace that is hard to describe if you have not watched it up close. A million open source projects are compounding on each other. New AI projects are compounding on each other. The people learning through that lens are compressing timelines that used to take years into weeks.
This is not a hot take about AI replacing engineers. It is about what has become possible for one person with good instincts and access to the right tools. Shipping a working product used to be gated by a long list of capabilities you had to build up sequentially. A lot of those gates are now flat.
The filter is pointed at the wrong thing
Most engineering loops are still grading juniors on the fundamentals of data structures and algorithms with the same intensity they always have. I am not going to argue that those fundamentals do not matter. They do.
But if the loop is only measuring that, it is not measuring what is actually changing in the field. A junior who is extraordinarily fluent with modern tools, who can move through ambiguity, who can compose working systems out of compounding open source, and who is already shipping products on their own time, is a different kind of signal. Grading them only on leetcode patterns is the wrong instrument for the moment.
The generation that stopped waiting
Engineering hiring has been volatile for a while. A whole generation of grads got filtered out for being too junior for entry level roles. A lot of them adapted. They stopped waiting to be hired and started building.
That is not a new story. What is new is how low the barrier has gotten. The engineers who in previous cycles would not have known how to start, fund, or build now can. The barrier to entry for starting a company used to mean chasing venture capital and clearing a long list of prerequisites. That is no longer true.
Some of the best talent right now is not as eager to break into FAANG. They are building tools, products, and companies. That has always existed at the edge. It is now happening at volume.
What the recruiting role should actually be
Recruiters have a role here that I do not see enough of in this market. Partnering with founders and hiring managers to make sure the company is not over-indexing on years of experience or pedigree. Not to lower the bar. To correct the instrument.
The conversation that matters is the one where the recruiter says, the strongest signal in this candidate is not the resume. It is what they built last month, how fast they are compounding, and how they think about systems now that the tooling has changed. That read is hard to surface if the rubric has not moved.
What this means for the next two to three years
I expect the shift to show up first in smaller companies that cannot afford to over-rotate on senior hires. They will start pulling heavily from the junior pool that has been quietly filtering itself out of the traditional path. The compounding effect will be faster than people expect, because this cohort has already been building for a year or two without anyone properly counting.
The companies that notice first will get a disproportionate share of a generation that otherwise would have gone elsewhere. The ones still reading years of experience as the cleanest proxy for capability are going to be the ones wondering why their pipeline looks weak.