Read Achilles
Human Capital & EdTech
~22 min read
TL;DR — Nobody is getting fired. Almost nobody is getting hired. That is the whole labor market right now: September added just 29,000 jobs while weekly jobless claims sat at 197,000, near a 57-year low. Wages rose just 3.0% over the year, the slowest pace since 2021, and continuing claims crept up to about 1.72 million: few people lose jobs, but those who do take longer to land. The freeze is the story — and inside the…
The opening paragraph, up front — the full story follows below.
Nobody is getting fired. Almost nobody is getting hired. That is the whole labor market right now: September added just 29,000 jobs while weekly jobless claims sat at 197,000, near a 57-year low. Wages rose just 3.0% over the year, the slowest pace since 2021, and continuing claims crept up to about 1.72 million: few people lose jobs, but those who do take longer to land. The freeze is the story — and inside the freeze, the one move that still pays is changing jobs, and the one promise that keeps breaking is training. Companies now offer AI courses to 58% of big-employer workers, up from 25% a year ago, yet 56% of those workers say they get no time at work to use them. The ladder is not being kicked away. It is being left out in the cold. For investors, the money question is just as sharp: training vendors are selling the cure for the freeze, staffing firms are priced for a thaw that has not arrived, and Coursera's Oct 29 report is the first number that separates the two stories.
What Needs Your Attention Today
Green means watching — we are tracking it and it looks normal. Amber means warning — it needs a closer look. Red means danger — time to act. The color never carries the meaning alone: every card says it in words.
The Picture in 60 Seconds
Think of the labor market as a frozen river. On the surface, nothing moves: no wave of firings, no surge of hiring. Underneath, the current still runs — toward health care and classrooms, away from finance and back-office work, and toward whoever is willing to change jobs. Where the analogy breaks: a river thaws on its own when spring comes. A labor market does not. It thaws when employers decide the uncertainty is over, and nothing in this week's data says they have decided that yet.
| Signal | Reading | Number | What it means |
|---|---|---|---|
| Hiring | Frozen | +29,000 | BLS September payrolls; weakest in months, far below expectations |
| Firing | Rare | 197,000 | Weekly new jobless claims; near 57-year lows |
| Jobless rate | Drifting up | 4.2% | Up from 4.1% in August; still historically low |
| Wages | Cooling | +3.0% y/y | Average hourly earnings, $37.81; slowest annual pace since 2021; workweek held at 34.4 hours |
| Moving premium | Alive | +2.9 pts | Changers' gross pay growth minus stayers' (ADP) |
| Training | Offered, not used | 58% / 56% | Offered AI training / report no time to use it (Workera) |
| Entry level | Split | +12% / 48% | PM postings up / managers who'd pick AI over a new grad |
The Big Picture: Low Hire, Low Fire — Still
Here's the argument: the labor market's defining feature is a freeze, not a collapse. Why it works this way: employers are hoarding the workers they have — firing is expensive when rehiring is uncertain — while refusing to add headcount until demand, rates, and costs clarify. The proof: the BLS September report (Oct 2) put payroll growth at 29,000, against expectations of roughly 84,000–88,000 in Bloomberg and Dow Jones surveys, and revised July and August down by 60,000 combined; the same week, new jobless claims fell to 197,000, below the 200,000 Reuters-polled expectation. What's already reflected in the numbers: the 4.2% jobless rate already carries the freeze — it rose only a tenth of a point despite hiring falling by two-thirds versus expectations, because so few people are being let go. What would change my mind: claims jumping above roughly 250,000 for several weeks running, or hiring plans surging back — Challenger's September hiring intentions (90,787, up from 12,325 in August but down 23% from a year ago, the weakest September since 2011) show neither yet.
ADP's private-sector count tells the same story from the other side of the ledger: 90,000 private jobs added in September, a rebound after a three-month slowdown, with base pay up 3.2% and gross pay up 4.7% year over year. ADP chief economist Nela Richardson called it "a strong report." Read both reports together and the picture sharpens: private payrolls are still growing, government and statistical revisions are dragging the headline down, and nobody — not the BLS, not ADP, not Challenger — sees a layoff wave. The freeze is a hiring freeze. That distinction matters, because hiring freezes end quietly, one job posting at a time, while layoff waves end loudly. If you are waiting for a headline that says "hiring is back," you will wait past the moment it actually came back.
The wage data sharpened the freeze. Average hourly earnings rose just 0.1% on the month and 3.0% over the year — the slowest annual pace since 2021 — to $37.81, and the average workweek held at 34.4 hours. Employers are not bidding up pay to hire, and they are not cutting hours either: they are simply standing pat. The household survey added a twist: employment rose 406,000 and the labor force grew 485,000, lifting participation to 61.8%. Part of the uptick in the jobless rate is people re-entering the market, not losing jobs — a softer reason for a worse-looking number. But the long-term picture is less kind: 1.9 million people have been out of work for 27 weeks or more, 27.1% of all unemployed.
The ADP–BLS gap needs a straight explanation, because it looks like one of them must be wrong. It is mostly a universe difference. ADP counts private payrolls from its own client records — 90,000 in September, the first acceleration since May, with August revised down to 36,000. The BLS private count was just 46,000; government payrolls fell 17,000, pulling the headline to 29,000. Same economy, different lenses: ADP sees private payrolls through one provider's books; the BLS surveys establishments across private and public employers and then revises. When the two disagree this much, treat the truth as somewhere between them — private hiring is growing, but slowly — and watch the revisions, which have run against optimism all year.
Claims deserve a second look, because the headline hides a split. The four-week average fell to 198,000, the lowest since early October 2022, and claims have stayed below 200,000 since the second week of September. But continuing claims — people still drawing benefits — edged up to about 1.72 million. Few people are losing jobs; those who do are taking longer to land. As Navy Federal chief economist Heather Long put it: “It’s still a ‘low-hire, low-fire’ job market.” That is the freeze from the job-seeker’s side.
Dispersion: The Split Inside the Freeze
The freeze is not evenly spread, and the unevenness is where the opportunity hides. ADP's September breakdown: education and health services added 55,000 jobs and leisure and hospitality added 22,000, while financial activities shed 16,000 and professional and business services lost 11,000. Goods-producing industries added 31,000, including 17,000 in manufacturing and 15,000 in construction. Translation: the economy is still hiring people who touch bodies, buildings, and food — and still shrinking the ranks of people who shuffle paper. The same pattern showed up in Challenger's year-to-date tally: announced cuts down 39% from last year overall, but concentrated in the white-collar layers that grew fastest during the pandemic.
The second split is the pay gap between movers and stayers. ADP's Pay Insights, built from payroll records covering more than 26 million private-sector workers, found total gross pay — everything in the paycheck, including bonuses and overtime, not just salary — up 7.3% for job-changers versus 4.4% for job-stayers: a 2.9-point premium for moving. On salary alone (base pay) the gap is narrower (4.8% vs. 3.0%), which means much of the mover advantage comes from bonuses, overtime, and hours, not just salary. The premium is real, but it is also a trapdoor: in a frozen market, the door you walk through may not open again behind you. A premium earned by leaving a stable job for a shaky one is not a premium at all.
The third split is the strangest. Product-manager postings are up 12% worldwide to a three-year high, according to hiring data reported by LinkedIn and Lenny's Newsletter — yet 48% of hiring managers in ResumeTemplates' survey of 1,000 managers say they would rather invest in AI tools than hire and train a 2026 graduate, and 55% say they have already shifted entry-level hiring budget to AI tools. Both can be true at once: employers are posting for experienced product leaders who can aim AI at real problems, while quietly defunding the junior rung that used to grow those leaders. The Verellax Index, which read 57,134 job ads across seven markets in Q3, found only 15.6% of entry-level roles even mention AI — and in Australia, just 4.9%. Employers are not writing AI into first jobs. They are writing first jobs out of the plan.
Challenger’s detail undercuts any simple “AI layoff wave” story. Technology was the one sector still cutting hard: 10,799 planned cuts in September, up 77% from August, and 165,925 for the year — up 54% on 2025 and 29% of all announced cuts. AI was the fifth-most-cited reason in September (3,961 cuts) yet remains the leading cited reason for the year (120,136, about 21%). Read that carefully: AI tops the year-to-date table mostly because every other reason fell faster, and Challenger records what employers cite, not verified replacement of workers by software. Food producers announced 7,326 cuts, the second straight month above 7,000, including 5,396 tied to Washington orchards and agricultural employers. Third-quarter cuts totaled 129,591, down 43% from the second quarter.
The holiday hiring signal is the softest in the report. Spirit Halloween and Michaels together announced 62,000 seasonal roles, against 100,800 last September. And the product-management paradox got sharper: the 2026 CPO Insights Report, surveying more than 1,500 chief product officers, concluded the PM role as it exists today will be obsolete by 2030 — even as postings sit at a three-year high. Surveys tracking AI job-loss anxiety found concern rising from 28% to 40% between 2024 and 2026, concentrated among mid-career white-collar workers. One career coach working with PMs put the average search at around six months, and warned that the freeze itself — months spent researching instead of shipping judgment — is what creates the real displacement risk. What the data does not settle is whether every posting is a live opening; treat posting counts as demand signals, not filled jobs.
What to Watch: The Trainers Move In — Without the Time
Here is the asymmetry hiding in plain sight: companies are buying the cure while skipping the dose. Workera's 2026 State of Skills Intelligence Report, based on a survey of 1,000 employees at organizations with 5,000 or more staff, found the share of companies offering AI-specific skills training rose from 25% to 58% in a year — while 56% of employees say they get no time at work to build the skills. The World Economic Forum's projection, cited in the same report, is that 59 of every 100 workers will need reskilling or upskilling by 2030, and 11 of them will not get it. Employers are moving — 77% plan to upskill in response to AI — but the bottleneck is no longer the course catalog. It is the calendar.
That reframes what to ask for. A course license is cheap; an hour a week on the clock is the actual benefit, and it is the one most workers are not getting. If your employer offers AI training, the question to ask your manager this week is not "which course?" but "which hour?" If you are a manager, the cheapest retention move in this market is protecting one hour a week for your team's learning — cheaper than a raise, and in a frozen market it may matter more. And if you are job-hunting under 30, read the Verellax finding as instructions: since barely one in seven entry-level ads mentions AI, the way to stand out is not another certificate — it is one piece of proof that you used AI on a real problem and checked the result yourself.
Workera’s detail shows three distinct breaks, not one. First, time: 84.3% of employees spend five hours a week or less on development, and 56.4% get no work hours at all for AI skills — even as 67.8% now use AI tools multiple times a week, up from 39.9% a year ago. Second, incentives: nearly 60% say AI skills carry equal weight to other skills, or no priority, in assignment and promotion decisions, and 43% cite a shortage of relevant materials. Third, visibility: 42.5% say they have been passed over, held back, or mis-rated because someone misjudged what they could do; only 26.6% get regular coaching from a manager and 21.5% get none; and 46.5% are building skills on tools their employer never provided — learning the company cannot see. One finding cuts against the automation narrative: 69.1% say a human manager would judge their skills more fairly than an AI tool. The bottleneck is not adoption. It is evidence: employers cannot see what their people can do, so they cannot reward it.
Coursera’s Oct 29 print now has a number to beat: the company guided third-quarter revenue of $364–372 million and adjusted EBITDA of $52–56 million, after folding in the Udemy combination. The question the print answers is whether enterprise demand for training holds while headcount is frozen — and whether the promised synergies are arriving on schedule. Watch enterprise demand and the tone on employer budgets, not just the headline beat or miss.
Value Creation vs. Value Capture
Value created: real, and it sits with the people who pair AI fluency with judgment. The mover premium (7.3% vs. 4.4%) is the market paying for demonstrated judgment in a new setting — and the education-and-health hiring surge (55,000 jobs) is the market paying for work that cannot be done by a model. Creation here is boring: proof of work, portable skills, hours actually spent learning.
Value captured: by the companies selling courses to the frozen. Training vendors capture the budgets that the freeze generates — 58% of large employers now buy AI training — while workers capture the value only if they get the time to convert courses into capability. Amazon captures a leaner Stores payroll; the fewer-than-1,000 workers cut in its latest round capture a pink slip during the company's own sale event. The strangest line of the week: employers say they would rather train AI than hire graduates, while posting more jobs for the experienced people who are supposed to have come from those graduate ranks. The ladder is not being climbed anymore. It is being billed by the hour.
What this means for you — your career. If you're employed: you are in the hoarded half, and that is leverage, not safety. Spend it asking for the one thing the data says works — hours on the clock for learning. If you're under 30 and job hunting: stop optimizing your résumé and start shipping proof. One portfolio piece that shows AI output plus your judgment beats ten certificates. And if you're a manager: the company that skips junior hiring to save money this year is the company hiring expensive strangers to be its mid-level managers in three.
What this means for you — your capital. The trainers' business model is a hedge on the cutters' business model, and hedges have prices. Coursera (NYSE: COUR) reports Oct 29 — that print is the tell, and the decision to make is before it, not after: if you hold training-vendor exposure, set your exit trigger now (for example, sell on any enterprise-demand miss rather than waiting to "see how it trades"). Here is the if/then checklist to run on the print: if enterprise demand (sales to big employers) holds up while headcount is frozen, training has become a must-have budget line and the training vendors earn their keep; if enterprise demand slips with headcount, training is still a nice-to-have and the vendors' growth story cracks. The sharper trade is two-sided: favor the platforms where training is built into the tools people already use every day (the software suites and workplace systems), and stay cautious on pure staffing firms whose business is charging per placed worker, since that model keeps shrinking as back-office work moves to fewer humans or none. Two more tripwires: if Thursday's claims print jumps toward 250,000, the freeze is thawing into something worse and safer, defensive holdings win; if Challenger's October tally shows hiring intentions recovering from their weakest September since 2011, the thaw is real and economy-sensitive stocks get interesting again. One more tripwire, with a number: ADP's mover premium stands at 2.9 points (7.3% vs. 4.4%). If that premium compresses toward 1 point or less in coming months, the reward for changing jobs is gone and the freeze has hardened into stagnation — favor stability over mobility in both career and portfolio decisions. If it holds near 3 points, mobility is still being paid for and selective job-switching remains the highest-return move available. Treat "AI-driven efficiency" on earnings calls as a layoff tell, and rising per-employee training time — not training spend — as the growth tell. Two more numbers to carry: the four-week claims average (198,000 — a rise toward 220,000 would signal the freeze cracking toward layoffs) and AI’s share of cited cuts (21% year-to-date — if September’s 9% pace holds, the ‘AI layoff wave’ story is overstated; if it climbs back above 20% monthly, it is not).
One last thing. The freeze will end. Every freeze does. The question is whether you spent it getting better or getting colder. The market is not going to announce the thaw. It will just start hiring again, one posting at a time — and the people who are ready will be the ones who used the quiet to build something provable.
What We Could Not Verify
Three limits. First, Challenger’s figures are announced plans, not confirmed separations or hires — a company can announce cuts it later reverses, and hiring plans are intentions. Second, the CPO Insights “obsolete by 2030” finding reaches us through a vendor release summarizing the survey, not the full report; treat it as a directional signal about executive expectations, not a forecast. Third, Amazon has not disclosed a country-by-country breakdown or severance terms for the latest round; the fewer-than-1,000 figure comes from a person familiar with the matter via Reuters, not a company statement. No figure in this issue rests on any single one of these.
Get the Daily Board in your inbox.
One email, every morning, with all 9 desks. Free. No spam, ever.
Home · Newsletters · Trend Indicators · MBTI · Education · Request · Privacy