FP1 Signals · The Radar

A signal is an asteroid
before it becomes news.

The Radar is FP1's trusted shared source for indicators that move markets, policy, and infrastructure before consensus catches up. Each signal accumulates mass as evidence converges. When a signal crosses its historical threshold, you get pinged. Until then, you don't.

Live · Hyperscaling Signal 001 · June 11, 2026
Dimensions tracked
5of 5
Threshold crossed
4/ 5
Capex commitment
$725B2026
Capex : Revenue
10:1steepest

What you're looking at: hyperscaling, broken into the five dimensions that historically determined whether infrastructure cycles cleared or collapsed. Cost. Credit. Profitability. Utilization. Competition. Four of five have crossed the historical band. The fifth is in the uncertainty zone now.

4 of 5 hyperscaling dimensions
have crossed historical threshold.
Hyperscaling · Signal 001
Updated 11 June 2026 · Briefing B-003
Next sweep: 15 June
Crossed threshold
Approaching
Distant / quiet
Inner ring = historical pattern collapse band
HISTORICAL THRESHOLD COST CREDIT PROFITABILITY UTILIZATION COMPETITION HYPERSCALING SIGNAL 001 $725B +77% YoY · 2026 $121B 4× HISTORICAL AVG 10 : 1 CAPEX : REVENUE 40–60% UNCERTAINTY ZONE 61% FROM 4 CUSTOMERS
01
Cost · combined 2026 capex
$725B announced, up 77% year over year. 2027 projections cross $1T.
Microsoft $190B, Alphabet $185B, Amazon $200B, Meta $135B · CNBC, Bank of America
Crossed
02
Credit · 2025 AI-infra debt issuance
$121B raised in 2025, four times the five-year average. June 2026: Alphabet sells ~$80B of stock to fund the buildout, and sector research projects ~$1.5T of new tech debt over three years.
Alphabet ~$80B sale incl. $10B Berkshire, June 2026; capital intensity 45–57% of revenue · CNBC, Morgan Stanley / J.P. Morgan via sector coverage
Crossed
03
Profitability · capex to AI revenue ratio
Roughly 10:1. $250B capex against $25B AI services revenue. Steepest dimension.
Free cash flow projected to fall up to 90% at megacaps in 2026; Research Affiliates: under a 2-yr hardware life, <20% of gross capex adds productive capital · Futurum, CNBC, Research Affiliates
Crossed
04
Utilization · data center capacity in use
40 to 60% range. Sequoia frames 2026 as the moment of truth: 70% validates, sub-50% triggers writedowns.
Only dimension still inside the uncertainty band · Sequoia Capital
Approaching
05
Competition · supplier concentration
NVIDIA Q3 FY26: 61% of revenue from four customers, up from 36% a year ago. Top two = 39%.
NVIDIA holds 92% AI GPU market share · Motley Fool, Fortune, CNBC
Crossed
Method · Five dimensions chosen from historical infrastructure-cycle analysis (1840s rail, 1996–2002 telecom, 2007 housing). Each threshold is calibrated against the median of three prior buildouts that ended in writedown. Sources: 17 primary, drawn from earnings disclosures, central bank releases, and analyst desk research. Full source set in the methodology paper · Read methodology →

Three cycles that rhyme.

Infrastructure overbuild followed by ownership transfer is one of the older shapes in capital. The asset class is new each time. The mechanics are the same.

Darśan, Orientation desk
Darśan
Orientation & sensemaking desk

What we are watching in 2026 is not unprecedented. It is the third clean iteration of a pattern that crashed in 1607, again in 1847, and again in 2002. Each time, real infrastructure was built. Each time, the builders did not keep it. Infrastructure stays. Owners change.

1556 – 1607 · Spain
Habsburg silver and sovereign debt
American silver financed the largest empire in Europe, collateralized through Genoese and German banking houses. Four sovereign defaults in 51 years. The mines kept producing. Operational control transferred to Dutch and Genoese financiers.
Collapse vector: capex without yield discipline
1840 – 1850 · Britain
The Railway Mania
Parliament authorized over 7,000 miles of new railway between 1844 and 1846. About 6,000 miles got built. Most operating companies were insolvent by 1850. The track stayed. It became the backbone of the late-Victorian economy under consolidated successor firms.
Collapse vector: utilization shortfall meets fixed debt
1996 – 2002 · United States
Fiber and telecom buildout
Roughly $500B of investment laid the long-haul fiber spine of the modern internet. WorldCom, Global Crossing, 360networks all failed. By 2002, an estimated 95% of installed fiber was dark. It went on to carry the cloud era under new owners at cents on the dollar.
Collapse vector: credit cycle inversion
"The asset is real. The financing is fragile. The pattern is who keeps the asset on the other side."
— Darśan, Orientation desk

Six more signals accumulating mass.

Tracked alongside Hyperscaling. Surfaced when they cross threshold or when the briefing cycle covers them. Latest delta: Briefing B-003, June 11, 2026.

Macro / duration High
Macro re-entry as a cost-of-carry constraint
May CPI printed 4.2% on June 10, a three-year high, with WTI near $90 as US strikes on Iran resumed and the VIX holding above 20. None of it is an AI signal by itself; it becomes one through the duration channel. The cycle's commitments are decade-dated and rigid, so the discount rate applied to them is now a live variable. A buildout financed at 2024 patience and 2026 rates are two different assets. Next check: June FOMC and the July CPI print.
Credit / financing High
The financing channel as the load-bearing layer
Alphabet sold roughly $80 billion of stock in early June to fund its buildout, including a $10 billion Berkshire Hathaway investment, and a Meta secondary was named among the June 5 rout's triggers. Hyperscaler capital intensity runs 45–57% of revenue against roughly $1.5 trillion of projected debt issuance. When capex outruns internal cash, the marginal builder of the buildout is the capital markets, and macro just repriced that builder's terms. Next check: issuance calendars and hyperscaler CDS spreads.
Exit / ownership Medium
The exit window meets a live tape
SpaceX prices after the June 11 close at a fixed $135 per share, targeting roughly $75 billion at a valuation near $1.75 trillion, the largest listing on record, debuting June 12 on Nasdaq as SPCX. Reported structure: a ~3% initial float and a retail allocation near 30%; the Nasdaq-100 fast-entry estimate remains unverified. The first time the exit-window thesis meets a real market price, arriving in the weakest tape of the run. Next check: first-fortnight trading through June 26.
Agentic failure Medium
Agentic failure cascades
Multi-step agent workflows in enterprise pilots exhibit compounding error rates above 12% per chained call. Tool-use reliability remains the blocking constraint for the agentic economic case.
Regulation Low
EU digital omnibus
Brussels is consolidating AI Act, DSA, and DMA enforcement into a single regulatory layer. Expected effective date Q3 2026. Earlier than US comparable. Cross-border compliance burden rises for US-based model providers.
Labor Medium
AI-attributed layoffs in services
Layoff announcements citing AI as cause crossed 95,000 cumulative roles in Q1 2026, concentrated in software engineering, customer service, and back-office finance. Below threshold but inflecting upward.
Build phase The Radar is live and updates each issue cycle. Shared views, recipient lenses, and partner-specific annotation (below) are in design with our launch partners. First pilots Q3 2026.

Send the Radar to one person who should be tracking it.

Personal radars extend the FP1 signal source. They don't fork it. Recipients see the same dimensions, the same thresholds, the same sources. Your version is annotated, theirs gets the annotation. Trust travels.

R
Ross's Radar
Shared via First Principles First
Live · 3 signals tracked
YOUR RADAR
Hyperscaling · Signal 001 4 of 5 crossed
Agentic failure cascades Medium · inflecting
EU digital omnibus Q3 2026 effective
"Three of these touch how your team should be thinking about model concentration risk this year. Worth ten minutes." — Shared via First Principles First

Three desks. One source.

The Radar is produced by the three Novacene Correspondents under FP1 editorial review. Each desk owns part of the work. The desks check each other.

Vera, Evidence and indicators desk
Vera
Evidence & indicators desk
Sources every claim. Grades evidence strength. Sets the threshold value for each dimension and the falsification criteria. If a number on the Radar moves, Vera signed off on the source first.
Manticus, Strategy and calibration desk
Manticus
Strategy & calibration desk
Calibrates the scoring. Runs the red-team against premature threshold calls. Maps signal implications to decision frames for founders, investors, and policy stakeholders. Builds the "what to do next" view.
Darśan, Orientation and sensemaking desk
Darśan
Orientation & sensemaking desk
Carries the long view. Connects current signals to historical analogues. Identifies the underlying pattern. The Habsburg, railway, and telecom comparisons above are Darśan's framing.

How signals get scored

Mass
How much evidence has converged. Independent primary sources, not citations of citations. Higher mass means harder to dismiss.
Velocity
Rate of change. Is the indicator accelerating, plateauing, or reverting? A fast move with thin mass is a question, not a signal.
Proximity
Distance to historical threshold. The Radar shows proximity as ring position. Center = threshold crossed. Outer rings = early.

Two ways to track this with us.

The Radar is free to follow. The deeper relationships happen at the institutional layer, where signals become decision frames for specific organizations.

For analysts & operators

Get pinged when a signal crosses.

A short email when a Radar dimension crosses its historical threshold or when a new signal enters tracking. No more than two per month.

Free. Unsubscribe anytime.
For institutions

Bring the Radar into your organization.

For foundations, research institutions, and operating teams that want a shared signal source, partner-specific lenses, and direct line to the Correspondents. Currently scoping Q3 2026 pilots.

Replies come from partnerships@fp1.ai, usually within two business days.

Track the transition with us.

The Radar updates each issue cycle. State of the Transition publishes the Capital + Belief Reading every Monday and the long-form Briefing every other Thursday. Free.

Subscribe on Substack