| ← New Money · Issue 009 · The Sunday Brief |
Sunday, May 3, 2026 |
|
◆ Issue 009 · May 3
The trillion-dollar
reallocation.
Where the money is moving — and what to do about your seat.
|
|
From the editor
Two structural reallocations are happening in parallel: the AI capex push pulling ~$420B/year into silicon and power, and the reshoring push moving $520B+ into US factories. Both are visible in 10-K filings. Neither is visible yet in most readers’ career plans. This issue is the bridge.
Reshored ops jobs paying $90/hr, the dollar exposure that’s about to bite per Bloomberg FX series, the four prompts every PM should template, the solo-skill earning table, and the actual wage-and-role transmission of AI capex.
|
The reshore wage premium: ops/maintenance at $90/hr.
Per BLS OES Q1 2026 cuts, semiconductor fabrication operators in Phoenix and Columbus clock $42–$58/hr base. Add overtime + retention bonuses + line-leader differentials and the working hourly comes to $78–$90/hr. Median 2026 hire age: 29. Median base education: 2-year associate, sometimes 4-yr in mechanical/electrical engineering. This is not the unionized factory wage of your parents’ era.
For office-floor adjacencies (operations PM, supply-chain analyst, EHS lead, automation engineer): $135–$185k base in those metros, plus $30–60k retention. Total package in low-COL state. The trade is: SF tech comp $300k+, Phoenix reshored ops $200k, COL difference $70k+/yr. Net of taxes, geographic comp delta > the brand-name premium.
What to do this week
Search LinkedIn for “operations program manager” at TSMC Arizona, Intel Ohio, Samsung Texas, Micron Idaho. Read three full JDs. Even if you’re not applying, the language tells you what skills the reshoring industry is bidding for.
|
The dollar’s 2026 vulnerability: hedge it like an offshore worker.
Per FRED DXY series, the dollar is down 9.4% from its November 2024 peak against a basket of trading-partner currencies. The catalyst stack lining up for further weakness: persistent twin deficits per CBO, Fed hold-then-cut paths, foreign central banks rebalancing reserves. Consensus 12-month range: −6% to −14% against EUR/JPY/CHF.
For readers paid in USD: every overseas trip and import you buy is about to cost ~10% more. For readers earning USD while spending EUR (Lisbon arbitrage from Issue 004): the dollar weakening cuts into the arbitrage. Practical hedges that aren’t exotic: 10–15% allocation to a hedged international equity ETF, holding 3–6 months expenses in EUR or CHF cash if you spend abroad, locking long-term euro-denominated obligations now if relevant.
What to do this week
Pull your portfolio. Calculate the % in unhedged USD assets. If >88%, you’re long the dollar by default. For under-35 readers with international consumption ahead, that’s probably the wrong tilt. Shift one ETF position to a hedged international equivalent this month.
|
The 4 prompts every PM should template once and never write again.
Templated Claude Projects, run weekly:
1 · COMPETITIVE TEAR
For [competitor], pull their last 3 release notes, blog posts, and price-page changes. Summarize: what they shipped, what they paused, what their pricing is signaling.
2 · USER RESEARCH SYNTHESIS
Read these 8 user interviews. Find 3 themes that appear in >5 of them. For each theme, give the verbatim quotes that prove it.
3 · 1-PAGER FROM A LOOM
Watch this 12-min Loom. Output a 1-page product brief: problem, user, hypothesis, success metric, three risks, what we’d build first.
4 · BOARD DECK RED TEAM
Here’s a draft board update. Be the most skeptical board member. List the 5 questions you’d ask. Mark each as “answerable now” or “requires new data.”
What to do this week
Save the four above as a Claude Project each. Time savings are real but the bigger win is the second-order one: the prompts force a recurring rigor on your week. Competitive teardowns become weekly, not quarterly. User-research synthesis becomes a habit, not a project.
|
The $4k/mo skill stack: solo earning rates, ranked.
Honest 2026 hourly-rate medians for solo operators with 5+ years of W2 experience, per public freelance-platform data, productized-consulting case studies, and reader replies. Ranked: (1) Series A–B fundraise prep $1,200–$2,400/hr (high-stakes, narrow buyer pool). (2) Equity-comp negotiation review $400–$900/hr. (3) AI-workflow buildout for ops teams $250–$500/hr. (4) SQL/dbt analytics-engineering retainer $200–$350/hr. (5) Technical writing / engineering blog ghostwriting $150–$280/hr.
The rate isn’t the bottleneck; distribution is. (4) and (5) are easiest to start because the cold-pitch buyer pool is wide. (1) and (2) are highest-paid because the buyer pool is narrow and the trust requirement is high. A reasonable solo plan: start at (4)–(5) for cashflow, route the trust capital toward (1)–(2) over 18 months.
What to do this week
Pick the line above closest to what you’d ship under your name today. Calculate: 12 hours/week × that rate = annual side income. Most readers will be surprised it clears $50k. The constraint is starting the conversation, not the rate.
|
AI capex: 30% of $420B is going to actual job categories. Which ones.
Of the $420B (Issue 001, Play 05), ~70% lands in silicon, datacenter build, and power. The remaining 30% — ~$126B — lands in actual labor-intensive categories. Per 10-K commentary and McKinsey AI org research: ML platform engineers (~$45B), AI-product PMs ($28B), MLOps/eval ($22B), domain-specific applied scientists ($18B), AI safety + governance ($13B).
The wage repricing inside these categories has already happened — ML platform staff IC at top labs clears $750k+. The categories that haven’t fully repriced: AI safety/governance and domain applied science. Both have hiring growth >40% YoY with median TC still under $400k. The premium will land in the next 18 months.
What to do this week
If you have any domain depth (healthcare, legal, finance, manufacturing) and basic ML literacy, the “applied scientist in domain X” track is where the next 18-month repricing is. Write a 1-page memo on the AI use case you’d ship in your domain. Publish it under your name.
|
◆ Chart of the Week
$126B of AI capex is labor. Here’s the split.
2026 AI labor spend, by category
| ML platform engineering | $45B |
| Domain applied scientists ★ | $18B |
| AI safety / governance ★ | $13B |
★ = not fully repriced yet. Where the operator window is.
Sources: SEC 10-K commentary, McKinsey AI org cuts. The bottom two lines haven’t fully repriced. That’s the operator opportunity.
|
◆ The Tape
$90/hr Reshored ops working rate |
−9.4% DXY YoY (Issue 002 macro) |
$126B AI capex labor share |
|
One favour before you go
Reply with one number: 01, 02, 03, 04, or 05 — which play you’d run first this week.
I read every reply. They shape what shows up in Issue 010.
|
|
Go run something. See you next Sunday.
— The Operator
still at McKinsey, still building
|
|
Know an operator who’d like this? Forward it →
© 2026 New Money
|
|