AKD TREASURY / FIELD INTELLIGENCE

Live Harvester

One verified Bitcoin Core source. A shared target, distributed work, mathematically verified returns, and learning that changes the next encounter.

Connecting to Core

Live Bitcoin Core · Historical crawler · SHA learner

Core validated

—

Bitcoin Core consensus blocks

Known headers

—

Not all fully validated yet

Research verified

—

Historical headers / SHA-256d

Crawler batches

—

Original Hub-owned batches

Core validation / initial block downloadAwaiting source

Core validation backlog—

Historical crawlerAwaiting event

SHA input/output witnessAwaiting first valid source

Source recovery pressureNone observed

Historical P2P research is independently PoW/ancestry tested, and remains distinct from fully Bitcoin Core-validated current mining work. A post-SHA 99.6% byte rejection is an output comparison, not a 99.6% saving in SHA computation. Browser and Motorola work still requires participant consent and a valid source job.

Original AQT · Bitwise transformer · One Ocean

Actual original Ocean source, double-SHA-256 mathematical trace, and source-to-teacher join. Live model training is only shown after a verified CHEREV 2 receipt.

Language observations

—

Original PSE encounters

SHA-256d traces

—

Original 192-round sources

Exact teacher joins

—

Source to round-state labels

Intelligent trials

—

Nonce + steering evidence

Bitwise model field640 header · 256 target · 192 SHA rounds

Teacher connectionAwaiting Ocean

CHEREV 2 AQT gradient and checkpointAwaiting native receipt

Observed original Ocean—

Learning evidence and verified model changes are distinct; nothing is promoted from a display.

Save your private account recovery token

This token gives access to your account. Store it securely; never post or share it.

Core network

—

Waiting for real template

Next block height

—

From getblocktemplate

Active epoch

—

Every new template cancels old work

Valid proof witnesses

Independent SHA-256d confirmation

Target doorway FULL 256-BIT

Block identity—

Compact nBits—

Bitcoin Core target · every digit must be satisfied

—

Merkle root—

Nonce frontier—

Local probes use actual current leased work; they never credit unverified claims.

Join compute starts opt-in, event-driven browser work. The current Core job and target are refreshed automatically. Stop anytime.

Shared Hub

Connected helpers0

Unique issued leases0

Rejected proofs0

Old-epoch returns0

Retained Core jobs0

Retained equation trials0

Verified hashes under full target0

Blocks accepted by Bitcoin Core0

Duplicate proof attempts0

A lease covers a disjoint nonce range for one job identity. On a new Core job, the old leases expire.

PSE learning · mathematical spine

Real Core jobs studied0

Equation candidate vs control—

The learner records both successes and failures. No improvement is promoted without held-out confirmation.

Gap to target

G = (T − H) modulo 2²⁵⁶, with the source's chosen H projection recorded explicitly.

Double-SHA relationship

80-byte header → two compression blocks → 32-byte digest → another compression. Track W3, carries, sigma, nonce and Merkle changes.

Learning by consequence

Compare directional candidates to equal-cost controls, retain what improves verified acceptance, and adjust bounded work distribution.

Hardware & connectivity

AIO CPU2 logical cores · measured

AIO GPUIntel Haswell GT1 · actual OpenCL SHA256d verified; mainnet compute waits for Core

Cherev nodeBitcoin Core 31.1 mainnet · IBD, not mining-ready

Phone helperAndroid Chrome SHA256d via USB · join/stop verified

Hub fleet presenceAwaiting joined participants

Pruned mainnet practice · verified evidence

Last studied header

—

P2P SHA256d, Core-anchored history

Comparisons per policy

—

Equal candidate budgets, new nonce values

GPU hashes CPU-checked

—

Historical mainnet experiment only

GPU target successes

—

Actual full 256-bit Bitcoin target

Practice evaluationAwaiting authenticated Ocean data

Current epoch equation to study nextAwaiting genuine trials

Paired control evidenceNo measured comparison yet

Equation admissionShadow learning only; proof required for network work

Real-chain SHA first-byte gate—

Gated candidate hashes—

Dynamic nBits / depth—

Both-and joined chain evidence—

Across-difficulty independent study—

Historical whole SHA-256d witness—

Nonce propagation into 192-round trace—

256/640-bit PSE equation experiments—

Exact target gap on historical winner—

Filter roleFirst-byte doorway after SHA256d; avoids deeper checks, not hash work

Historical header studies remain separate from an active Bitcoin Core mining template. No predictive SHA256d advantage is claimed without independent held-out evidence.

Intelligent Hashing · Motorola + WIMELE

One original ROL Hub dispatches fresh verified mainnet historical SHA-256d evidence to the real Motorola ARM64, checks every result independently on the AIO, and changes the next bounded phone assignment. Research is distinct from a current mining job.

Phone / current workflowReading current source

Verified historical phone hashes—

Measured Motorola kernel / full transport—

Learned change to next actual batch—

Original Ocean accepted learned events—

Current test pressure—

Observed events

Bitcoin Core jobs, joined-device readiness, and independently verified learning arrive through one Hub event stream.