Sixfold Skyreading
A working framework for seeing institutional events coming, before the press tells you it was inevitable.
In one sentence
Sixfold Skyreading is a deliberate counter-bias — a process for re-pointing the analyst’s gaze, on a known cadence, at six independent classes of signal that institutions emit before something surprising happens — designed against both the single-arrow prior that committing to a question prevents seeing, and the information-overload failure where the signal is present but the analyst is too snowed-in to see it.
The working metaphor is Converging Clouds, with two bears underneath it.
Two bears, one discipline
Consider two distinct ways of failing to see what’s in front of you.
The first bear is upstream of the question. The analyst, the journalist, the AI user has a prior — a thing they cannot stop thinking about — and that prior shapes the question they ask. The question shapes the answer they receive. They never see anything that wasn’t already in the question. Single-arrow priors driving single-arrow questions.
The second bear is downstream, out in the world. The signals exist. The data is being collected. There is a polar bear standing fifty meters away, in plain sight. But the analyst is in a blizzard — snowflakes falling so heavily that every snowflake demands the same attention as every other, and the bear, despite being large and present and visible-in-principle, becomes functionally invisible. Inattentional blindness, scaled to civilization.
The empirical anchor for the second bear is Christopher Chabris and Daniel Simons’ Gorillas in Our Midst (1999). Subjects are told to count basketball passes between players in white shirts. A person in a gorilla suit walks onto the court, beats their chest, and walks off — for nine seconds. About half of viewers, focused on the counting task, do not see the gorilla at all. Awareness, it turns out, is not a passive thing that registers what is in front of you. It is a budget. (In a 2013 follow-up, even radiologists searching chest scans for lung nodules failed to see a gorilla photoshopped into the image 83% of the time.)
This second failure is not because the data is missing. The data is abundant. Carnivore — the FBI’s email-surveillance program of the late 1990s and early 2000s — was reading internet traffic at the backbone. Its successors collect more, faster, with better fusion. Customs filings, eBay markups, secondary-market price moves, social media heat, satellite imagery, leaked Slack channels, podcast tone shifts, hobbyist forums — the snowstorm of available signal in 2026 is so heavy that an attentive observer can drown in it. The bear is fifty meters away. We just cannot see it.
Sixfold Skyreading is the counter-bias against both bears, simultaneously. A fixed list prevents the first bear: you cannot ask a single-arrow question because the framework requires six readings. A scheduled, repeating practice prevents the second bear: you re-look at the same six places on a known cadence, regardless of what is loud that week.
The point is not the cleverness of the analyst. The point is the discipline of the cadence.
Why the framework exists
Every dominant prediction technology we have is structurally optimized for one cause per story:
- The newspaper article (one headline, one villain, one arrow).
- The sell-side analyst note (one revised price target, one driver).
- The AI chatbot (one fluent paragraph, one confident probability).
- The classroom case study (one teachable lesson, one named decision).
Reality is not arranged this way. Real institutional events almost always emerge from the simultaneous lighting-up of three to six independent vectors that converge in a narrow window and produce a surface event the press then narrates with a single arrow.
The convergence is systematically hidden, not accidentally hidden. Companies suppress it because it implies internal disagreement. Journalists compress it because deadlines demand a single arrow. AI assistants reproduce the compression because they were trained on the suppressed-and-compressed corpus. The discipline of seeing convergence is therefore a deliberate counter-bias, not a natural human capacity.
The six vectors (v0.1)
A skyreading walks this fixed list before issuing any prediction about an institution we are watching:
| # | Vector | What to watch | Where it leaks |
|---|---|---|---|
| 1 | Personal/biographical of key principals | Age, health, wealth events, family transitions, public-tone shifts | Profiles, podcasts, social media |
| 2 | Internal accountability pressure | Quiet reorgs, missed targets, named-and-shamed business units, comp disclosures | Proxy filings, 10-K language shifts, mid-level departures |
| 3 | Successor readiness | Who is being given visible scope? Who presents on calls? Who does the press? | Earnings call rosters, conference assignments, internal promotions |
| 4 | Strategic disagreement | Public statements that contradict each other, sudden product-cadence changes | Patent filings, R&D disclosures, conflicting interviews |
| 5 | External shock exposure | Geopolitics, supply concentration, regulatory venues, currency exposure | Trade data, customs filings, central bank actions |
| 6 | Demand-side surprise | Sudden sellouts, secondary-market price moves, community forum heat | eBay, Reddit, niche forums, hobbyist communities, field observation |
The detection rule (v0.1, will be calibrated against filed predictions over time):
- Three or more vectors lit up in a rolling 60-day window → treat the institution as convergence-vulnerable. Begin filing dated, falsifiable predictions (Oracle Bones).
- Four or more lit up → expect a surprise event within 90 days.
A skyreading is run on a fixed cadence (weekly for the watchlist, monthly for the broader sky). The cadence is the discipline. The cadence is what defeats the bear in the blizzard.
A working example
Apple, April 11–30, 2026. All six vectors lit up in three weeks: Mac Mini and Mac Studio supply collapsed (vector 5 + vector 6); Tim Cook announced his retirement effective September 1, with John Ternus (SVP Hardware Engineering) as successor (vectors 1, 2, 3); Apple’s AI position remained visibly behind, prompting strategic-disagreement signals from inside the company (vector 4); an Iran strike layered geopolitical pressure onto the supply chain (vector 5).
The financial press told this as four separate stories with four single arrows. A skyreading on Apple — even one performed by a careful outsider with no privileged access — would have flagged convergence-vulnerability by mid-April and a likely surface event by the end of the month. The signals were public. The framework to read them across silos was not.
(Worked out in detail in an internal learning memo, May 3, 2026.)
Why it matters in a teaching context
Strategy students are taught case studies in single-arrow form. They graduate believing real institutions move that way. They do not. The 494BI capstone, in particular, is the right venue to teach skyreading — not as an academic theory, but as the discipline of refusing the single-arrow narrative when it is offered, regardless of whether the source is the Wall Street Journal, a McKinsey deck, or an AI chatbot.
The pairing with Single-Arrow Fallacy is intentional: the disease and the cure each get a name, so the diagnosis can travel.
Trade-offs and warnings
- Skyreading is slower, less marketable, and less confident than single-arrow prediction. It will lose sales contests. It is intended to win predictions over time.
- Information asymmetry is real. Many vectors require access we do not have (boardroom conversations, internal reorgs, classified intelligence). A careful outsider can detect convergence, but should be modest about predictive claims until the framework has a track record.
- False positives matter. What is the base rate of three-vector convergences that do not produce a surprise event? Without that, the framework will over-predict. The Oracle Bones discipline (write down, date, grade) is the only honest way to find out.
- The framework is v0.1. Treat the rules as scaffolding, not law. The vectors and the thresholds will be revised as the bones get scored. The “Sixfold” in the name commits us to the current vector count; if a future version lands on five or seven, we rename. (We have already accepted, elsewhere in the system, that zhengming — the rectification of names — happens when the architecture moves under us.)
See also
- Single-Arrow Fallacy — the bias this entry counters
- Grey Swans — what happens when skyreading is missed
- Oracle Bones — the prediction-filing practice
- Aunties — operational oversight; skyreading is the epistemic counterpart
- Space Cowboy — the user class who most needs this discipline