Evidence, confirmed by a human,
becomes structural intelligence.
One illustrative Space, followed end to end: evidence → candidate observation → human confirmation → signals → metrics → phase → direction. Every figure below is illustrative and was hand-checked against PENOCH's actual formulas.
Watchlist: NVDA, AMD, TSM, ASML, AVGO, ARM, MU, SMH, SOXX (the “AI Infrastructure” preset).
AI-proposed observations, pending review. Confirm what's accurate, discard what isn't — only confirmed observations affect PENOCH state (Signals, Metrics, Phase, Direction below).
4 pending review · illustrative controls, not wired to a live engine
A widely-read semiconductor newsletter flagged early signs of order pushouts from a mid-tier cloud customer, contrasting with hyperscaler capex guidance.
A dissenting data point against the dominant “capex is durable” narrative — worth confirming to track whether it's isolated or the start of a pattern.
A secondary foundry reported in-line but not beat results, with cautious language on next-quarter utilization.
Weaker than the sector's recent beat-and-raise pattern — a possible early crack in the earnings-quality signal, not yet corroborated elsewhere.
Some institutional 13F filings show trimmed positions in mega-cap AI infrastructure names during the most recent quarter.
Runs against observation #2 (capital rotating in) — could indicate profit-taking by a subset of holders rather than a reversal.
A leading memory manufacturer confirmed high-bandwidth memory (HBM) capacity is sold out through the next two quarters.
Reinforces the existing supply-constraint read (observation #4) from a different point in the supply chain.
Leading AI accelerator supplier beat revenue and margin expectations for the third consecutive quarter, citing sustained hyperscaler capex.
Earnings Quality +4 · Demand Strength +2
Institutional flows continue rotating from broad-market index funds into semiconductor and AI-infrastructure names.
Capital Concentration +3 · Capital Availability −2
Cloud hyperscalers raised full-year capital expenditure guidance for the second time this year, citing AI compute demand.
Demand Strength +1.2
Advanced packaging capacity (CoWoS) remains fully booked through next year, constraining accelerator shipments.
Supply Constraints +3
Forward valuation multiples on AI infrastructure names have expanded well above their 5-year average.
Valuation Expansion +0.9 · Supply Constraints +0.6
Sell-side commentary has shifted from “AI capex is a bubble risk” to “AI capex is structurally durable” over the past two quarters.
Sentiment Divergence +1.8 · Earnings Quality −0.6
The AI infrastructure basket has outperformed the broader semiconductor index by a wide margin over the past 20 trading days.
Capital Concentration +2
8 structural signals. Each score is its baseline plus the sum of impacts from every confirmed observation, weighted by that observation's confidence.
5 metrics, each the equal-weighted average of the signals mapped to it.
Driven by: Capital Concentration
Driven by: Demand Strength, Supply Constraints, Valuation Expansion
Driven by: Earnings Quality, Sentiment Divergence
Driven by: Innovation Velocity, Capital Availability
Driven by: Sentiment Divergence, Capital Availability
Confidence: Medium
Rules are checked in order; the first fully-satisfied one wins. With Concentration 75, Pressure 54, Coherence 56, Permission 51: Zombie, Reset, and Stressed all fail (Pressure isn't high enough). Consolidating matches — Concentration clears its 60 floor by 15 points, and Pressure clears its 55 ceiling by only 1 point. Because the tightest margin (1) is far under the 10-point bar for a High-confidence read, the phase confidence is Medium, not High — this is a real, rule-derived margin, not a placeholder.
Since day 1 of this Space · gated on ≥5 confirmed observations (7 confirmed here) — with fewer than 5, PENOCH would honestly report “Insufficient Data” rather than guess.
52 → 54 (+2)
54 → 56 (+2)
Both moved by the same +2, but the algorithm reads this as Slightly Deteriorating, not Stable or Improving: Pressure's +2 clears the 1-point “slight” threshold, while Coherence's +2 falls short of the 3-point threshold needed to register as Improving. Rising Pressure is treated as the dominant signal until Coherence's own improvement is large enough to outweigh it — a real asymmetry in the production logic, not a narrative choice made for this sample.
This Space's structural state can be snapshotted into SEUS RT and contribute to Evidence Confidence. It does not alter SEUS Opportunity/Risk/Positioning scores or the verdict — see the SEUS sample →