Onpode
Cover art for AI has reached half of all jobs—yet hiring postings hit 12-year lows

AI has reached half of all jobs—yet hiring postings hit 12-year lows

November 16, 2026 · 7 min

Maya Chen & Dr. Nathan Hayes

Maya, quick question before we start — you job-hunted in tech before, right? Walk me through that memory for a second. Ha — yeah, once. The job boards felt endless. Almost too much to scroll through. Why? Now imagine the same board, but with roughly a third of the listings gone — and every remaining…

As of November 14, 2026, Hacker News's monthly "Who Is Hiring?" thread recorded 244 job postings, down 37% from 387 at the same point in 2025. Every thread since July 2026 has landed between 224 and 255 postings; August's 224 was the fewest since January 2014. The 2026 yearly average of 287 postings per month is the lowest annual average since 2013.

0:006:43
Get the next episode on Artificial Intelligence →

Follow it free — new episodes land in your feed.

Or make your own — any topic, in minutes

More Onpode episodes on Artificial Intelligence →

About this episode

As of November 14, 2026, Hacker News's monthly "Who Is Hiring?" thread recorded 244 job postings, down 37% from 387 at the same point in 2025. Every thread since July 2026 has landed between 224 and 255 postings; August's 224 was the fewest since January 2014. The 2026 yearly average of 287 postings per month is the lowest annual average since 2013.

Grounded in 6 sources
Nobel Economist Daron Acemoglu on Why Democrats Should be Worried ↗ · bloomberg.com
How AI may reshape career pathways to better jobs | Brookings ↗ · brookings.edu
"Who is Hiring?" postings stuck near 12-year lows as AI reaches half of all jobs | Hacker News ↗ · news.ycombinator.com
Daron Acemoglu on What Three Years of AI Labor Data Finally Prove – Authority Journal ↗ · authority-journal.com
IT Hiring Trends 2026 | CRB Workforce ↗ · crbworkforce.com
Goldman Sachs AI Job Loss Forecast: 9% of Workers Explained | explainx.ai Blog | explainx.ai ↗ · explainx.ai
Read transcript

Dr. Nathan Hayes: Maya, quick question before we start — you job-hunted in tech before, right? Walk me through that memory for a second.

Maya Chen: Ha — yeah, once. The job boards felt endless. Almost too much to scroll through. Why?

Dr. Nathan Hayes: Now imagine the same board, but with roughly a third of the listings gone — and every remaining one demands a skill that didn't exist professionally three years ago. That's what hntrends.com is showing for the Hacker News 'Who Is Hiring?' thread right now. 244 postings as of November 14, 2026. Down 37% year-over-year from 387.

Maya Chen: Wait — 37% in one year?

Dr. Nathan Hayes: One year. And August hit 224 postings — fewest in a single month since January 2014. Now, the part that doesn't resolve cleanly: 51.6% of November's surviving postings mention AI. That share crossed 50% for the first time in June 2026, up from roughly 30% two years prior. The board is contracting and intensifying simultaneously.

Maya Chen: So it's not that hiring collapsed because AI interest collapsed — it's almost the opposite.

Dr. Nathan Hayes: Exactly the paradox. And PwC's 2026 Global AI Jobs Barometer widens this — roughly half of all globally advertised job categories carry substantial AI task exposure. So what I want to work out today is whether the Hacker News collapse is a tech-sector recession, an AI displacement mechanism, or some third thing entirely.

Maya Chen: But third thing — that's the one I can't let go of. Because 287 postings a month as the 2026 yearly average, lowest since 2013 — that's not a blip. That's a trend line. So is this actually a tech contraction, or are we watching something narrower than that?

Dr. Nathan Hayes: Narrower. Importantly, the S&P Global employment index puts the net global impact at negative five points over the prior year — with a marginal further decline forecast. So there's a real signal there. But Hacker News is one hiring thread, in one tech-adjacent community. What it tracks well is startup and mid-cap tech appetite. What it doesn't track is the rest of the labor market.

Maya Chen: So the macro's not screaming.

Dr. Nathan Hayes: Not yet. And this is where I need to slow down on PwC's language — because the 2026 Global AI Jobs Barometer says roughly half of advertised job categories have substantial AI task exposure, and that gets read as 'half of jobs are at risk.' PwC explicitly does not say that. Exposure measures which tasks AI could theoretically touch. Actual displacement depends on adoption costs, regulation, whether new demand fills the gap. Those are completely different questions. Now, the forecaster range on actual displacement is — genuinely wide. Goldman's Joseph Briggs, in a note reported October 22, 2026, estimated that even with a 15% productivity lift, unemployment rises maybe 0.5 percentage points, temporarily, then resolves. That's the optimistic anchor. The other pole — Anthropic's CEO — has warned of 10 to 20% unemployment. Same underlying data, wildly different models.

Maya Chen: Ten to twenty — that's not a rounding error from 0.5. That's a different world.

Dr. Nathan Hayes: It is. And I'd be misleading you if I said we know which model is right. What I can say is the aggregate macro figures — right now — are not validating the pessimistic pole. But here's what aggregate figures hide, and this is actually where it gets worse: picture a software engineering grad who finished in December 2025, sending applications in March 2026. The macro unemployment rate looks fine. Her cohort employment data does not. We'll get to that specific collapse in a minute — Brynjolfsson's numbers on 22 to 25-year-olds are where the floor actually drops out.

Maya Chen: So macro calm and cohort crisis — those can coexist.

Dr. Nathan Hayes: They coexist, and the Brynjolfsson data is where that becomes undeniable. Stanford Digital Economy Lab — Erik Brynjolfsson's team — found a 16% employment decline for 22 to 25-year-olds specifically in AI-exposed U.S. occupations since November 2022. Not the whole labor market. That cohort. That slice.

Maya Chen: Sixteen percent since 2022. That's — hmm. That's not a blip, that's a sustained drain.

Dr. Nathan Hayes: And Goldman Sachs puts Big Tech new-graduate hiring down 25% between 2023 and 2024. So picture someone — spring 2026 grad, CS degree — applying for roles that, two years prior, explicitly said 'new grads welcome.' Now the same posting says 'two to three years of AI integration experience required.' That's not an upgraded preference. That's a structurally different role masquerading as the same title.

Maya Chen: So the ladder didn't disappear — the bottom rung just got yanked up.

Dr. Nathan Hayes: That's the mechanism exactly. And Robert Half data puts CS graduate unemployment at 6.1% — well above overall market stability. Now, the part I actually find counterintuitive: Robert Half also finds 29% of companies are quietly rehiring after AI-driven cuts. So it's not pure contraction. It's churn — but the floor on reentry is higher. The on-ramp itself moved.

Maya Chen: Wait — rehiring after cuts? So some of those roles are coming back, just... requiring someone who already knows AI integration?

Dr. Nathan Hayes: Right. And this is where Acemoglu's firm-level data matters — his research, now elevated after the 2024 Nobel, shows top-quintile deep integrator firms captured a 31% productivity premium. Bottom three quintiles? One to four percent. So most firms aren't even getting the gains that would justify raising the bar. They're raising it anyway, I mean — that's the part that doesn't fully resolve. They're pattern-matching on what the top firms did without the actual infrastructure.

Maya Chen: So you're locked out of the rung not because AI actually replaced you — but because companies think it did, or think it will, and they're hiring accordingly.

Dr. Nathan Hayes: That's the uncomfortable part, yes. And I don't know that the data resolves it cleanly — what Acemoglu's Nobel work actually shows is that the gains are real but so concentrated that most firms are, in a sense, hiring for a future they haven't built yet. So the floor holds — 224 to 255 postings every month since July, that narrow band — but it doesn't rise. That's the signal worth watching. Not whether the unemployment rate spikes. Whether that floor ever lifts.

Maya Chen: Mm. You started this by asking me to picture a job board with a third of the listings gone. I think — yeah, I think that's still the right image. Except now I'd add: the ones left are asking for skills the person who needs that first rung doesn't have yet. And official statistics won't name that until it's already set.

Dr. Nathan Hayes: Quietly structural. That's an honest place for us to land — thanks for thinking through it with me.