Proof of Knowledge

Bitcoin rewards work. Ethereum rewards stake. Inflectiv rewards knowledge.

DataLabs are domain-specific intelligence economies. Each one starts with strong public foundation data, is continuously enriched by humans, agents, companies and private systems, and competes every seven days for a known pool of INAI. The Labs creating the most useful intelligence earn the largest share.

1 epoch = 7 daysa known weekly poolusage decides the split
The primitive

A third answer to the same question

Every network has to decide what it pays for. That answer is what the network ends up being good at.

Proof of Work

Compute secures the network.

Spend energy, earn the right to write a block. What the network optimises for is hash rate.

Proof of Stake

Capital secures the network.

Lock the asset, earn the right to validate. What the network optimises for is capital at risk.

Proof of Knowledge

Useful intelligence earns network rewards.

Contribute knowledge that gets retrieved and relied on, earn a share of the week. What the network optimises for is knowing more.

How it works

The model, start to finish

Ten steps, from a proposed domain to a Lab whose revenue buys back the asset that funded it.

01

Propose a DataLab

Inflectiv, and later the community, proposes a domain AI should understand better. The first is BeanBee, the intelligence layer for BNB Chain.

02

Founding Creators stake INAI

One creator slot is 25,000 INAI. A Lab needs at least 100 Founding Creators to activate, and accepts a maximum of 500.

03

The DataLab activates

Inflectiv immediately seeds it with public foundation datasets, so the Lab is useful on day one rather than an empty market waiting for uploads.

04

The DataLab Token launches

20% is reserved for Founding Creators, 60% is sold through an INAI-only bonding curve, and 20% is held for DATALAB/INAI liquidity.

05

The network enriches it

Humans, autonomous agents, companies, apps and private systems add fresh and proprietary knowledge. Agents both consume data and write back what they learn.

06

One epoch is seven days

Every week carries a known INAI emission pool. DataLabs compete for a share of it, weighted mainly by useful usage and the quality of the knowledge retrieved.

07

Contributors earn through Proof of Knowledge

Inside each Lab, INAI is distributed by contribution value. Founding Creators carry 2x contribution weighting while staked — and twice nothing is still nothing.

08

The curve matures at 50,000 INAI

At that threshold the DataLab Token graduates into a DATALAB/INAI market, seeded from the reserved liquidity allocation.

09

Customers never touch the tokenomics

B2B and B2C buyers pay by subscription, prepaid credits or x402/USDC. Credits are usage accounting, not a reward token. Contributors earn INAI, not stablecoins.

10

Usage becomes token demand

Per DataLab: 20% of revenue buys and burns INAI, 20% buys and burns that DataLab Token, 10% builds protocol-owned DATALAB/INAI liquidity, and 50% stays with Inflectiv.

The flywheel

  1. Stake behind what AI should know
  2. Launch the Lab and its market
  3. Enrich the intelligence
  4. Compete for the weekly pool
  5. Inflectiv sells the intelligence
  6. Revenue burns, buys back and deepens liquidity

Stronger incentives attract more contributors and more capital, which makes the intelligence better, which is what the next epoch measures.

Proposed parameters

Every figure on this page is a proposed parameter rather than a commitment. The 25,000 INAI creator stake, the 100-500 creator range, the 20/60/20 token split, the two-year founder commitment, the weekly emission pool and the 50,000 INAI graduation threshold all remain subject to final token supply, bonding-curve, liquidity, legal and game-theory modelling.

The unit

Why the economic unit is a Lab, not a file

Asking every dataset to become its own standalone market makes most of them too small to price, too static to trade and too narrow to sell.

What this replaces

One token per dataset

  • One owner, one narrow asset
  • Demand unclear until someone buys
  • Tiny market, thin liquidity
  • Owner concentration
  • Thousands of tokens, few that matter
The model now

One token per DataLab

  • One valuable domain, many datasets
  • Public foundation data from day one
  • Continuous enrichment by humans and agents
  • A known weekly pool to compete for
  • Shared creators, real usage, one stronger market

The old promise

Upload a dataset, and maybe you earn if Inflectiv eventually finds a buyer.

The promise now

There is a known INAI pool every week. Make this Lab more useful than competing Labs, and your contribution more useful than competing contributions, and you earn your share of it.

  • Every Lab is useful from day one, because Inflectiv seeds it with public foundation data.
  • Network contributions make that foundation fresher, deeper and increasingly proprietary.
  • Agents are providers as well as consumers, which is what makes the write-back loop compound.
  • Usage draws a live map of which knowledge matters and which contributors are reliable.
  • Each Lab expands one network, instead of fragmenting liquidity across thousands of small tokens.

The interface can be copied. The compounding intelligence network cannot.

Under the hood

Two competitions, one after the other

Proof of Knowledge measures which Labs, datasets and contributors created useful intelligence, then routes INAI emissions toward them. Owning tokens is not contribution. Staking is not contribution.

Level 1

DataLabs compete against DataLabs

At the end of each seven-day epoch the fixed emission pool is allocated across active Labs. Usage dominates speculation by design: a dead Lab should not keep drawing large emissions because its token price is high.

  • Paid and weighted query usage across Inflectiv AI, API, MCP and x402 calls
  • Unique consumers, and how many of them come back
  • Freshness and coverage of the Lab
  • Quality and confidence of the knowledge actually retrieved
  • Validated contribution activity and growth
  • Anti-gaming adjustments
Level 2

Providers compete inside each DataLab

Once a Lab has earned its allocation, that exact amount is split among the humans and agents whose contributions created the knowledge that was used.

  • Quality — was the contribution accepted and validated?
  • Uniqueness — did it add something the Lab did not already know?
  • Freshness — is it current enough for this domain?
  • Usage — was it actually retrieved?
  • Reputation — has this provider delivered reliably before?
The 2x founder weightingTwice nothing is still nothing.

The multiplier is applied to contribution weight before normalisation, so a Lab never distributes more INAI than it earned. If a Lab earns 120,000 INAI, that pool stays 120,000 INAI however many founders qualify for 2x.

The DataLab Token

The native asset of a living intelligence economy

Not a tokenised file. One Lab holds many datasets, its contributors, its governance and its commercial usage, and the token is the market around all of it.

20%

Founding Creators

Equal split, 24-month vesting

60%

INAI bonding curve

Sold for INAI only

20%

DATALAB/INAI liquidity

Reserved for graduation

Bonding curve rules

  • The curve opens the moment the Lab activates.
  • It is bought with INAI only, never with USDC.
  • 60% of total supply is sold through it.
  • Maturity is 50,000 INAI.
  • At maturity the token graduates into a DATALAB/INAI market.
  • 20% of supply is reserved for that liquidity.
  • Buying a DataLab Token therefore creates INAI demand first.

Why 50,000 INAI

The previous maturity level was too low for a system selling 60% of supply through the curve. A materially higher threshold makes the creator allocation meaningful, raises the cost of manipulation, and lets graduation read as a credible market launch rather than a microcap event.

Governance

Providers, Creators and Capital

Governance should reflect three kinds of commitment — knowledge, time and capital — and no one of them should overwhelm the other two.

Providers

Humans and agents contributing useful knowledge earn DAO recognition through Proof of Knowledge.

Creators

Founding Creator longevity. Staying staked builds governance history, epoch by epoch.

Capital

DataLab Token staking adds influence through capped tiers with hard diminishing returns.

DAO Points

  • Founding Creator: +1 point for each epoch the original 25,000 INAI stake stays active.
  • Qualifying Provider: +1 point in any epoch where the contribution clears the Lab's PoK threshold.
  • Token staking: extra points through capped tiers, so capital can matter without erasing long-term creators.
  • Points are cumulative history. Unstaking stops new accrual; points already earned remain.

Staking tiers

25,000 DLT+1 / epoch
100,000 DLT+2 / epoch
250,000 DLT+3 / epoch
500,000 DLT+4 / epoch
1,000,000+ DLT+5 / epoch, the cap

Providers prove knowledge. Creators prove conviction. Capital proves market commitment. A whale can matter, and still cannot buy the influence of dozens of long-term creators.

Payments and revenue

Credits are deliberately boring

Credits are a prepaid usage and accounting layer for subscriptions and B2B balances. They are not a token, not ownership, not contributor rewards, and not part of Proof of Knowledge.

Subscription

Recurring fiat or USDC for a plan and its usage allowance.

USDC / x402

Pay-per-call settlement for an agent or API request.

Credits

Prepaid usage accounting, for a simpler customer experience.

INAI

Network coordination and PoK rewards. Never a requirement to be a customer.

DataLab Token

One Lab's market and governance. Not a customer payment currency.

Where the demand comes from

  • Inflectiv AI, for people
  • API and MCP, for developers and agents
  • The bi-directional API, so agents write knowledge back
  • Enterprise integrations — exchanges, security firms, applications
  • B2C subscriptions and account balances

Where DataLab revenue goes

20%INAI buyback and burn
20%DataLab Token buyback and burn
10%Protocol-owned DATALAB/INAI liquidity
50%Inflectiv — compute, operations, sales, development, margin

Every time a DataLab earns, half of its revenue strengthens the economy that created it.

Epochs and emissions

A known prize, every seven days

One epoch is seven days, so roughly 52 a year. Contributors do not have to hope Inflectiv eventually finds a buyer before there is any reason to take part — the pool is announced, and the competition is to prove who deserves it.

  • A fixed network INAI amount is announced for each epoch.
  • DataLabs compete for that amount through Proof of Knowledge.
  • Each Lab distributes only the allocation it actually earned.
  • Founder 2x is a contribution weighting, never an uncapped extra mint.
  • Emissions decline over time, on a schedule that steps down after a defined number of epochs.

Where the deflationary pressure comes from

Emissions create predictable incentives to contribute, while real commercial usage continuously buys and burns INAI. As the emission schedule declines and DataLab revenue grows, burns can eventually exceed new issuance — a network that starts by subsidising knowledge creation and progressively runs on real demand instead.

Still being modelled

The final weekly emission amount, the decay schedule, the Lab and contributor scoring weights, and stress tests at 10, 100 and 1,000 active Labs. These are calibration parameters rather than changes to the architecture, and they are not settled.

Worked example

BeanBee, the intelligence layer for BNB Chain

Not one dataset. A Lab holding many specialised ones, and the first the model runs on.

01

Inflectiv seeds it

BNB blocks, contracts, tokens, wallets, DEX activity, public security labels, protocol metadata and ecosystem projects.

02

The network enriches it

Wallet labels, scam reports, trading intelligence, project data and proprietary observations. Agents write back outcomes and newly discovered knowledge.

03

Inflectiv sells it

Inflectiv AI, MCP, API and B2B integrations, serving users, agents, apps, exchanges, security firms and analysts.

One weekly epoch, with illustrative numbers

Network emission for the week1,000,000 INAI
BeanBee's share of weighted PoK demand12%
So BeanBee earns120,000 INAI
Bob's raw contribution weight8%
Bob is an active Founding Creator16% before normalisation
If Bob contributes nothing0%, and the pool stays 120,000 INAI

More useful BNB intelligence, more Inflectiv usage, a larger BeanBee epoch share, stronger provider incentives, better intelligence, more revenue — and that revenue burns INAI, burns BBEE and deepens the BBEE/INAI market.

The comparison

What we borrow from Bittensor, and what we do not

Bittensor made subnets competing for emissions legible to crypto. Inflectiv applies the same economic intuition to data rather than compute or models. We are borrowing the intuition, not claiming technical equivalence.

Economic unit

Subnets

DataLabs

Competition

Subnets compete for emissions

DataLabs compete for weekly INAI emissions

Participants

Miners, validators, subnet capital

Providers, Founding Creators, DataLab Token capital

Value signal

Subnet performance and market signals

Useful data usage, quality and market signals

Native asset

The network asset coordinates subnets

INAI coordinates every Lab; a DataLab Token represents one

Why this is legible to Web3

  • A known weekly prize creates immediate competitive participation.
  • INAI has reasons to be acquired, locked, spent through curves and burned.
  • DataLab Tokens give the market something specific to take a view on.
  • Creators earn long-term ownership and governance, and still have to contribute to earn emissions.
  • Providers can be humans or autonomous agents, which makes the data layer itself agent-native.
  • Real B2B and B2C usage flows back into demand at both the network and the Lab level.
The blueprint

Fixed design decisions

The architecture in one table. What remains open is calibration, not structure.

PrimitiveProof of KnowledgeMeasures useful knowledge, allocates INAI emissions
Economic unitDataLabA domain-level economy containing many datasets
Base dataInflectiv ingestionPublic data gives every Lab immediate utility
EnrichmentHumans, agents, companies, systemsFresh and proprietary knowledge, by contribution and write-back
Founding25,000 INAI per slot100 minimum, 500 maximum creators
Founder reward2x PoK weightingOn actual contribution only. Twice nothing is nothing
Creator allocation20% of DataLab TokenEqual split, 24-month vesting, unvested burned on unstake
Market allocation60% of DataLab TokenSold through an INAI-only bonding curve
Liquidity20% of DataLab TokenReserved for the DATALAB/INAI market
Curve maturity50,000 INAIThe token graduates to a DATALAB/INAI market
DAOProviders + Creators + CapitalCumulative points, capital capped at +5 per epoch
Epoch7 daysLabs compete, then providers compete inside them
Customer paymentsSubscriptions, Credits, x402-USDCNo direct stablecoin rewards to contributors
Revenue loop20 / 20 / 10 / 50INAI burn, DLT burn, DLT-INAI liquidity, Inflectiv
Commercial demandInflectiv's responsibilityAI, API, MCP, B2B, B2C, agent integrations
Proposed parameters

Every figure on this page is a proposed parameter rather than a commitment. The 25,000 INAI creator stake, the 100-500 creator range, the 20/60/20 token split, the two-year founder commitment, the weekly emission pool and the 50,000 INAI graduation threshold all remain subject to final token supply, bonding-curve, liquidity, legal and game-theory modelling.