The Data Forecast
Every morning at 04:00. No institution. No pay. Better than Nexus.
The forecast appears on G Nook terminal screens in amber text. It appears in chalk shorthand on Lamplighter junction walls. It moves through the Undervolt by word of mouth starting at 04:30, so that by 06:00 every informed Dregs resident knows what kind of day they are walking into. No corporation produces it. No institution maintains it. No one is paid.
The inputs are publicly available Grid load data, Counted member observations, Lamplighter harmonic measurements, and the particular instinct of people who have lived in the thermal shadow long enough to feel a server farm spin up through the soles of their feet. Pencil-47 — the forecast's primary author — compiles the daily report using colored pencils and hand-drawn matrices. Pencil-47's predictions outperform Nexus's internal load-balancing projections by a margin that has widened each of the last three years. Nexus's sensor grid cost an estimated 4.2 billion credits to deploy.
The Sprawl's most accurate model of its own electromagnetic environment is maintained by a person the formal economy has officially deprecated, distributed through chalk marks on concrete, and funded by nobody. The forecast works because no corporate model has ever attempted to measure what it measures. Nexus models server load as a compute variable. The forecast models server load as weather — something that falls on people, that they breathe and walk through and make daily decisions inside of.
Dregs residents opted into the Sprawl's computational infrastructure without opting into its costs. The server farms that run their Second Mind subscriptions and employer systems generate heat, harmonic interference, and electromagnetic fog that falls unevenly — concentrated in interstitial zones where land is cheap and regulations are historical curiosities. An entire population navigating daily life through conditions their employers generate, their infrastructure propagates, and their economy has never priced. The forecast is not a solution to this. It is a map.
Technical Brief — The Vocabulary
The forecast uses terms that didn't exist before the Cascade. Nexus has its own vocabulary for the same phenomena. The vocabularies do not overlap, because they are not measuring the same thing.
Load Weather
Anticipated electromagnetic conditions derived from server farm activity schedules. "Heavy load expected S4-D 1400–2200" means interface lag, power surges, and a temperature spike in the Dregs starting mid-afternoon. Corporate employees receive this information automatically through employer-provided environmental monitoring. Dregs residents receive it from chalk on a wall. The information is identical. The delivery infrastructure is not.
Thermal Index
Predicted temperature differential between corporate and interstitial zones, expressed in degrees. A thermal index of +6 means the Dregs will run six degrees warmer than Nexus Central. High-index days make the Undervolt dangerous for extended habitation. Low-index days remove the Dregs' only free heating. The preferred range is +2 to +4. Residents call anything above +7 a "cooker." They do not have a name for days below +1. On those days the priority is not vocabulary.
Fog Probability
Likelihood of sustained electromagnetic interference, expressed as a percentage and estimated duration. "Fog 70% / 8hr" means plan for a day when neural interfaces misfire, Second Mind subscriptions drop packets, and forced-focus contracts count employer-generated lag against personal performance scores. Workers on forced-focus contracts check fog probability before anything else. A high-fog sick day costs a shift's wages. A high-fog work day costs more. The forecast gives them the numbers. The decision is still theirs.
Cascade Risk
Probability of harmonic cascade on a 1–5 scale. Level 4 and above are not published. The reasoning is circular and correct: publishing a Level 4 prediction would cause mass interface disconnection as residents panic, and mass disconnection would itself trigger the cascade the prediction warned about. The forecast's most important function is the one it cannot perform in public. Pencil-47 has recorded seven Level 4 readings in the last two years. All seven were communicated privately to the Lamplighters, who adjusted junction harmonics without public explanation. The public noticed nothing. That was the point.
The Forecasters
The Counted members who contribute observation data walk their routes before dawn — noting temperature differentials at junction points, the pitch of transformer hum, the feel of the air where corporate zones bleed heat into interstitial corridors. The Lamplighters provide harmonic measurements from the junction network, taken on analog instruments because digital sensors in the Dregs are subject to the same interference they'd be measuring. (This is not a workaround. This is the only method that works.)
The observational methodology the Counted uses requires months of training to internalize. The harmonic readings the Lamplighters provide require calibration expertise that Nexus technicians would recognize as professional-grade if they encountered it in a context they respected. None of it is compensated. The forecast exists in the economy the way air exists in the economy — essential, unpriced, and therefore invisible to anyone measuring value in credits.
Nexus measures the same electromagnetic environment with a sensor grid spanning forty-seven sectors, AI-driven analytics pipelines, and executive dashboards that cost more per quarter than the Dregs' entire informal economy generates per year. The sensor grid produces a less accurate picture of the compute climate. The discrepancy is not mysterious. It reflects what each system was built to observe: Nexus measures what affects server output. The forecast measures what affects the people the servers are running on top of. These are different questions, and different questions produce different data.
Implications
The forecast's accuracy is a structural problem for anyone who would prefer the data not exist. If Pencil-47's predictions are more accurate than Nexus's internal models, that means either Nexus's sensor grid is optimized for the wrong measurement (it is), or the forecast contains data sources Nexus cannot access (it does — human observation in zones where corporate monitoring infrastructure is sparse). Both conclusions are uncomfortable in different ways.
The Power Auction uses forecast data to set daily bidding strategy. The Dream Exchange delays settlement when the forecast predicts drought conditions. Patience Cross checks fog probability before deciding whether to open the noodle shop on high-interference days. The forecast has become load-bearing infrastructure for the informal economy — which means Pencil-47 has become load-bearing infrastructure for the informal economy, and Pencil-47 is one person with pencils and no institutional protection.
The Sprawl's most consequential meteorological service has no redundancy, no backup author, and no succession plan. Pencil-47 has not identified this as a problem worth solving. The analyst who compiled this file is not certain whether that represents extraordinary confidence or something quieter.
Related Systems
| Entity | Relationship |
|---|---|
| Pencil-47 | Primary author. The colored-pencil matrices are the forecast's analytical engine. The forecast is, in material terms, Pencil-47. |
| Data Weather | The phenomenon the forecast predicts — electromagnetic conditions produced by the Sprawl's computational infrastructure falling on the people beneath it. |
| The Counted | Member observation data feeds the forecast's ground-truth layer. The Counted's routes exist for mutual aid; the forecast data is a byproduct of that work. |
| The Lamplighters | Harmonic measurements from the junction network. They also act on Level 4+ data without explanation. The public relationship is infrastructure inputs. The private relationship is emergency coordination. |
| The Power Auction | Forecast determines daily bidding strategy. High-load forecasts change what power access is worth and who bids for it. |
| The Dream Exchange | Delays settlement when forecast predicts drought. The Exchange's operations are timed around predicted interference. |
| Forced-Focus Contracts | Workers use fog probability to calculate daily sick-day decisions. The forecast is the only tool available for this calculation. |
▲ Unverified Intelligence
- At least one Nexus infrastructure division has been cross-referencing internal load projections against the public forecast for the past eighteen months. Whether this is competitive benchmarking or something more operational is not confirmed.
- Pencil-47 has declined three separate offers to share methodology in exchange for undisclosed compensation. The offers are described as coming from "institutional parties." No source has named the institution.
- The seven Level 4 cascade readings are not documented anywhere the public can verify them. Lamplighter coordination logs for those dates show "routine harmonic maintenance." The Lamplighters' definition of routine may require revision.
- The forecast's distribution via chalk shorthand uses a notation system the Lamplighters developed internally. The system was not formally shared with the forecast. Pencil-47 apparently adapted it independently. The Lamplighters have no recorded explanation for how this occurred.