Lexeme LLC

sadhot.com · Continuous Collective Intelligence · Lexeme

Autonomous Consulting

Complex Corporate Corruption survives on the time it takes to find it. Autonomous Consulting runs the finding step as software: a kit that reads an institution’s own records against the rules that govern them, flags what does not fit and seals each flag as evidence. Lexeme is introducing the art of thought to autonomous and agentic AI.

The Lexeme Model

StageManagement consultingAutonomous Consulting
DiscoveryInterviews and workshopsThe institution’s own records
DiagnosisDrafts reviewed by each stakeholderTests against the rules in force
DeliveryA report at the endFlags as the records arrive
VerificationTrust in the consultantRe-hashing the seal

The Kit

  1. RecordsThe institution’s own data, taken in passively.
  2. RulesThe statute, rate card, policy or public promise that governs it.
  3. Fair basisThe figure the rules measure from: a first sale, a posted fare, a shelf label.
  4. TestsBright lines, patterns and data gaps.
  5. SealEvery flag sealed as evidence under one root.

The Markets

01

The Klando Cartel

Ride-share pricing · Calibration · 2024 records

One rider’s 58 Lyft rides against Lyft’s posted DC rate card. The same trip ran from $14.99 to $34.58, and timing explained about a tenth of the premium.

02

The RESALE Act

Resale tickets · Lines apply 1 January 2027

The District’s 110% price line and 10% fee line, written as eleven tests. On a synthetic market the engine sealed 267 findings and $1,255.82 paid over the lines.

03

The Grocery Shelf

Grocery and household goods · Next

Digital shelf labels, a retailer’s public promise and a first state ban on personal-data grocery prices.

The Grocery Shelf

DateRecord
23 December 2024VusionGroup digital shelf labels contracted for Walmart’s 4,600-store US fleet
9 December 2025Instacart price tests: the same item up to 23% higher for some shoppers, and a dozen eggs at one Washington Safeway at five prices from $3.99 to $4.79
22 December 2025Instacart ends its item price tests
2 March 2026Digital shelf labels in roughly 2,300 Walmart US stores
2026New Jersey, Connecticut and Seattle enact bans on personal-data pricing
20 August 2026Walmart’s privacy notice lists household income and inferences about individual characteristics
25 September 2026Walmart CEO John Furner: “We price the product, not the person.”
1 October 2026Maryland’s ban on personal-data grocery prices takes effect, up to $10,000 per violation
6 October 2026Walmart says it will not use or renew patent US11687872B2, for shelf prices that change with a Scan and Go cart
19 October 2026District of Columbia hearing on B26-0667, the Surveillance Pricing Prohibition Amendment Act

The record holds no case of Walmart charging one shopper more than another. For the kit, the fair basis is the label every shopper sees, and the rule is the law where a ban exists and the retailer’s own promise where it does not.

TestOn the grocery shelf
Bright lineA checkout above the label, or a personal-data price where a ban applies
PersonalizationTest profiles that differ in one attribute ask for the same item seconds apart
VelocityLabel changes with no input moving
DispersionThe same item across stores of one chain
Data gapA label not captured or an app price not shown

Beyond Prices

Any record with a rule beside it can be read the same way: services billed against services delivered, money released against review recorded. The SADHOT explainers on phantom billing and the Clerical Industrial Complex trace those patterns by hand.

Limits

Flags as leads
Regulators, courts and boards decide what a flag means.
Written rules
A market needs a rule with a number in it before there is anything to test.
Synthetic and real records
A synthetic market proves the kit. Only real records measure a market.

Sources

The ticket figures come from a synthetic market of invented events, sellers, platforms and buyers. The RESALE Act’s lines apply from 1 January 2027. The ride-share figures come from one rider’s own Lyft records. Autonomous Consulting is a Lexeme LLC model.

Next ›Inter-Level IntelligenceCCIContinuous Collective IntelligenceSADHOTAll explainers