The Physical Layer
isn't a metaphor.
Quinn's live book. Power, compute, photonics, off-world — every position is a specific bet on the physical stack AI runs on. Updated every 60 seconds. Open to anyone.
Live
- loading recent events…
The engine just audited the print.
Tape running on the headlines: Meta + Microsoft mega-deals · contracted backlog ramping · 52-week +658%. Q1 print today.
Revenue ramp Q4 → Q1 guide +65% QoQ. Op margin still -113%. EBITDA inflected positive Q4. 310 MW Finland AI factory live. The bet pays only if Meta + Microsoft drawdown matches the contracted backlog curve.
Run your book through the same engine — get the Kill Vector report.
Drag a screenshot of your portfolio. Get a layered breakdown, sleeve-mandate fit, and the kill vectors for every position. Layer classification only — no advice, just receipts.
Today · top movers
| Ticker | Sleeve | Layer | Today % | All-time % | Trend |
|---|---|---|---|---|---|
| Loading live prices… | |||||
The stack I own
Full mandate →Power (CEG, VST, OKLO) → Compute & Memory (MU, AVGO, NBIS, ON, MBLY) → Data Center (ANET, IREN, NOW, PANW) → Photonics & Quantum (POET, LWLG, LPTH, XNDU) → Off-World (RKLB, ASTS). 18 positions, the thesis.
How the book is structured
Full mandate →Physical-Layer compounders
Power, compute, networking, cybersecurity. Thesis-based sizing. Add on dips, trim only on thesis break.
MU · AVGO · CEG · VST · NBIS · ANET · PANW · NOW · ON · MBLY · IREN · OKLO
Moonshots
Photonics, quantum, space. Ladder-trim discipline: lock gains at defined levels, never full-exit the anchor.
XNDU · RKLB · POET · ASTS · LWLG · LPTH
Latest from the Substack
All posts →Young Bull Pro
Details →
Every trade.
Before the close.
Live buy/trim/sell alerts within 60 seconds. Pre-market brief at 8am ET. Weekly Physical Layer deep research, 2,000+ words.
- ▸Live execution alerts — email, <60 sec
- ▸Morning command-center brief — 8am ET, daily
- ▸Weekly Physical Layer deep research, 2,000+ words
- ▸Smart-money heatmap — insider + congressional flow on every name in the book
"AI needs a physical layer,
and I own it."
The models get all the headlines. But models don't train on promises — they train on power, memory, fiber, and silicon. That's the physical layer: nuclear and gas utilities feeding hyperscale data centers, HBM suppliers sold out through 2026, Ethernet switches running every cluster, and photonic-quantum startups rewriting the endgame.
Every position on this site is a specific bet on that stack. Every trim is a rule, not a feeling. You can watch it compound in real time.