This fall, for its second cohort, the software company Palantir will pay a group of teenagers who never went to college to do the work its engineers do.
The program is called the Meritocracy Fellowship. It runs four months, it pays a real salary, and it is open only to high school graduates who are not enrolled in any university.[1]
The inaugural class, last year, took 22 fellows out of more than 500 applicants and paid them $5,400 a month.[1][2] The job posting did not bury the pitch. It read, “Skip the debt. Skip the indoctrination.”[1] The pitch was to earn the Palantir credential instead.
The company’s chief executive, Alex Karp, has been blunter still. On an earnings call he said that once you are inside Palantir, no one cares whether you went to Harvard or to no college at all, and that Palantir itself is by far the best credential in tech.[3]
Part of that pitch speaks to a culture-war audience, and that is the part that got the attention. The indoctrination line drew sharp reaction online, and Palantir defended it. I am setting that argument aside, because it is not what makes the move interesting.
The interesting part sits underneath the noise.
Palantir did not stop screening people. It stopped trusting the degree to do the screening for it, and started looking at what each applicant could actually do. That is not a story about college being worthless. It is a story about a shortcut that stopped working, and what a sharp buyer does when that happens.
Everyone announced this. Almost no one did it.
For a few years now, dropping the degree requirement has been the fashionable thing for an employer to announce. Google, IBM, Walmart, Accenture, Delta, all of them put out word that you no longer need a four-year degree to apply.
It made for good press.
Then researchers checked whether it changed who actually got hired.
In 2024, the Burning Glass Institute and Harvard Business School studied what happened after companies stripped degree requirements from their postings. The postings changed enormously. The hiring barely moved.
They found the great skills-based hiring revolution had created new opportunity for, in their words, not even one in 700 hires.[4] Almost all of the real change came from about a third of the firms. The rest announced and did nothing.
That is what makes Palantir worth watching. Most companies dropped the degree on paper and kept hiring the same people. Palantir actually changed who it hired. That is rare enough to be a data point rather than a slogan.
Why a smart employer would do this
To see why, look at what the degree used to do for the person doing the hiring.
For decades a degree was a reliable shortcut.
An employer could read it as a quick promise that someone could show up, finish something hard, and do skilled work, and be right often enough that the shortcut paid. That is what a good signal does. It saves you the work of judging each person one at a time.
The shortcut is getting less reliable.
By early 2026, the unemployment rate for recent college graduates had climbed to about 5.7 percent.[5] A degree still beats no degree at the same age. Young workers without one sit higher, north of 7 percent. But look at what the cushion is worth now. Recent grads used to sit well below the rate for the whole workforce, around 4 percent. Today they sit above it. And about 42 percent of recent graduates are underemployed, working jobs that never required the diploma at all.[5] I traced that oversupply, and how the country keeps blaming the wrong people for it, in 25 Million in Crisis.
Here is the part that matters, and it is easy to miss. None of that means a degree is a bad deal. It means the results have spread out.
A degree from the right school in the right field, bought at the right price, is still one of the best investments a young person can make.
A degree from the wrong one, in an oversupplied field, at full price on loans, is now a likely loss.
The average sits between those two, and it has slipped. So the label by itself, just holding a degree, tells an employer less than it used to. The information did not vanish. It moved from the diploma to the specifics underneath it.
Then there is what is happening to the bottom rung.
Stanford researchers, using actual payroll records, found that since late 2022 workers aged 22 to 25 in the jobs most exposed to artificial intelligence saw their employment fall about 16 percent relative to peers in jobs AI barely touches.[6] The entry-level job a degree used to buy is exactly the job AI is starting to affect. So the diploma points at a narrower target than it used to, at the same time the target is moving.
When a shortcut stops predicting the thing you bought it to predict, you stop leaning on the shortcut. You do not stop hiring. You start looking closer. That is not ideology. That is arithmetic.
This is already the consensus, quietly
If Palantir were alone, you could call it a stunt. It is not alone. The shift that earns headlines in the private sector has already happened, with far less noise, in government.
Maryland dropped the four-year-degree requirement for thousands of state jobs in 2022, the first state to do it. Utah followed and removed the requirement from 98 percent of its classified state positions. Pennsylvania opened about 65,000 jobs, roughly 92 percent of its state workforce, to applicants without a degree.[7] These were governments run by both parties, and they did it for the plainest reason there is. A blunt requirement was screening out capable people for no good reason.
So the Palantir fellowship is not the leading edge of a fringe idea. It is a loud version of something a lot of serious institutions have already worked out.
It did not drop the bar. It moved it.
Here is where the honest version parts company with the marketing.
Palantir did not abolish the filter. It swapped a blunt one for a sharper one. To be considered for that inaugural class, you needed an SAT score of at least 1460 or an ACT of at least 33, a result in the top few percent in the country.[1][2] The company that says it does not care about pedigree still screened hard for ability. It traded the degree as the gate for a test and a tournament.
That is the accurate way to read all of this. The degree is not going away. It is losing its spot as the one filter everyone has to pass through. Reading what a person can actually do is harder than reading the name of their school, which is exactly why most companies only pretended to do it. The degree lasted as long as it did because it was a cheap, reliable proxy. It is losing ground now because it became an expensive one that points to a wider and wider range of outcomes.
The same discipline, from the other side
I spend most of my time looking at this from the student’s side of the table, running the return on a degree as a financial investment. The lesson there is the same one Palantir just acted on. Do not trust the label. Look at the specific case.
When Palantir decides a degree is no longer enough to tell it who to hire, it is making the same move a family should make when it decides whether a degree is worth the price. Both are refusing to let a category stand in for the person. The employer stopped asking did you graduate and started asking what can you do. The family should stop asking does college pay and start asking does this college, in this field, at this price, pay for this student.
There are still plenty of great bets on that board. That is the whole point. The trouble is the good ones are hard to spot. They sit buried under a pile of signals that have nothing to do with whether a school actually delivers, the prestige, the pretty campus, the famous name, the high price tag that gets mistaken for quality. The averages do not say walk away. They say the easy rule of thumb stopped working, so do the homework the rule used to do for you.
There is a hopeful reading here too. A hiring system that looks at what you can do, rather than what you bought, is a door for exactly the people the old shortcut shut out. The talented kid who could not afford four years, or could not finish, or never tested into the right zip code. Done honestly, looking past the label widens the gate. The danger is the fake version, the press release with nothing behind it, which lets a company look open while it changes nothing. I make the fuller case for finding and growing that overlooked talent in America’s Talent Crisis.
The signal worth keeping
So set the culture-war pitch aside. It is the least interesting thing Palantir is doing. The interesting thing is the discipline underneath it. A sharp buyer looked at a signal it had trusted for years, saw that it had stopped sorting cleanly, and went to the trouble of judging each person on the merits.
That is the move worth copying, on both sides of the table. The degree is not going to vanish, and for the right student it is still one of the best investments around. But it is no longer a number you can read off the top of a resume and trust. The people who do best from here, families and employers both, will be the ones willing to look past the label at the thing it was always supposed to stand for. If you are weighing that bet for yourself, that is the work. You can start by running the return on your actual school and major at collegeroi.org.
Reference Sources
- Palantir Technologies. “Meritocracy Fellowship.” Palantir Careers, recruiting for the Fall 2026 cohort, https://www.palantir.com/careers/meritocracy-fellowship/. Accessed 15 June 2026.
- “Palantir Says Colleges Are No Longer a Reliable Training Ground as It Hires 22 High School Students for Its Fellowship.” Fortune, 5 Nov. 2025, https://fortune.com/2025/11/05/palantir-colleges-no-longer-reliable-training-ground-gen-z-hire-22-high-school-students-fellowship. Accessed 15 June 2026.
- Karp, Alexander. Palantir Technologies Second-Quarter 2025 Earnings Call, 4 Aug. 2025.
- The Burning Glass Institute and Harvard Business School Project on Managing the Future of Work. “Skills-Based Hiring: The Long Road from Pronouncements to Practice.” Feb. 2024, https://www.burningglassinstitute.org/research/skills-based-hiring-2024. Accessed 15 June 2026.
- Federal Reserve Bank of New York. “The Labor Market for Recent College Graduates.” Federal Reserve Bank of New York, 2026, https://www.newyorkfed.org/research/college-labor-market. Accessed 15 June 2026.
- Brynjolfsson, Erik, Bharat Chandar, and Ruyu Chen. “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence.” Stanford Digital Economy Lab, 13 Nov. 2025, https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/. Accessed 15 June 2026.
- State of Maryland (2022), State of Utah (2022), and Commonwealth of Pennsylvania (2023) executive actions removing four-year-degree requirements from most state classified positions.
This piece is a companion to my work on college return on investment, including College Is a Bet, which works out the odds from the student’s side. You can run the return on any school and major at collegeroi.org.