Is the subnet producing intelligence that people or systems repeatedly choose to use?
AlphaSign in
Proprietary Bittensor investment intelligence
Found Durable AI demand?
Trade with discipline.
Our thesis: Bittensor turns useful machine intelligence into investable markets. Centurion Alpha uses competing AI researchers to identify the subnets earning real demand, test what could make the thesis wrong, and propose risk-sized trades for human approval.
AI assists the decision. It does not hold keys, bypass policy, or guarantee a profitable outcome.
Invest behind useful intelligence—not a ticker, narrative, or model answer alone.
Bittensor makes AI markets.
We look for the ones that matter.
Bittensor subnets compete to produce digital intelligence. Our opportunity is to identify where useful output, developer adoption, sustainable economics, and market liquidity reinforce one another before that quality is fully reflected in price.
We do not assume every subnet—or every AI model—deserves capital. A position must survive evidence review, an opposing case, liquidity checks, explicit invalidation conditions, and portfolio risk limits.
Can demand, incentives, and emissions support value beyond temporary attention?
Can we enter, size, and exit without ignoring liquidity, slippage, or concentration?
Did the thesis produce repeatable net results after costs, not merely a persuasive story?
Models advise.
Evidence decides.
K3 is a preferred candidate, pending verification of its canonical Bittensor subnet identity, API contract, availability, and performance on Centurion's trading research benchmark.
Bittensor-powered models are preferred when measured quality, cost, latency, and reliability meet the task.Research
AI agents scan subnet fundamentals, market structure, liquidity, catalysts, and disconfirming evidence.
Challenge
Independent models build bull, bear, and risk cases so a fluent answer never becomes automatic conviction.
Verify
Claims retain sources, timestamps, confidence, and invalidation rules, then get scored against outcomes.
Constrain
AI can recommend a trade; portfolio limits, approval gates, and external signing control whether it moves.
No AI provider receives trading authority. Recommendations remain attributable, reviewable, policy-bounded, and separate from signing.
One decision package. Produced competitively. Scored against reality.
Market noise is abundant.
Conviction is earned.
Centurion Alpha replaces opaque advice with a market of competing intelligence producers—each measured on what happens after the recommendation.
Every packet carries its evidence, confidence, downside, invalidation conditions, and permitted action. Polished stories do not outrank useful results.
Observe
Chain, market, product, developer, and customer evidence enters the network.
Compete
Specialist miners produce forecasts, research, risk, and allocation signals.
Validate
Validators score calibration, evidence, originality, latency, and outcomes.
Decide
Consensus becomes a portfolio proposal constrained by the user's policy.
Machine-readable signals with provenance, not anonymous hot takes.
Evidence in.
Decisions out.
The Investment Intelligence Packet is signed, timestamped, reproducible, and scored later against realized outcomes.
Packet protocol v1.0Subnet rankings
Dynamic, evidence-backed rankings across utility, liquidity, adoption, and risk.
Portfolio policy
Risk-controlled TAO allocations shaped by your limits, horizon, and convictions.
Catalyst signals
Fresh alerts for revenue, integrations, listings, governance, and on-chain change.
Risk intelligence
Drawdown, concentration, liquidity, and invalidation signals before action.
Forecasts without evidence are opinions. Evidence without a scored outcome is research. Centurion Alpha joins the two.
Recommendations can move only as far as the declared rules permit.
Intelligence proposes.
Policy constrains.Humans approve.
Capital stays protected.Transcript baseline
A labeled benchmark—not investment advice. Subnet identity, liquidity, and pricing must be verified dynamically.
Customer revenue is a possible result of demonstrated success—not an assumption built into the launch.
Prove the system.
Then earn the right to offer it.
Paper performance can authorize only a controlled live pilot using the Centurion's capital. It cannot authorize customer billing.
Premium access remains locked until a pre-registered live window produces positive net returns after costs, stays inside its risk limits, reconciles to independent records, and passes independent review.
Paper proof
Settle deterministic strategies, costs, drawdown, and benchmarks before any capital is exposed.
Centurion live pilot
Use only capped Centurion capital with human approval, external signing, and complete reconciliation.
Verified financial success
Show positive net live returns after costs across a frozen 90-day minimum window and independent review.
Premium may follow
Only demonstrated Centurion success can unlock consideration of fixed-price premium customer access.
Backtests, paper gains, short windows, and selected winning trades do not satisfy the commercial gate. Historical success would support consideration of premium access; it would never guarantee future returns.
TARGET 14 / CONDITIONAL COMMERCIAL STAGE
Centurion success first.
Premium may follow.
If the Centurion first demonstrates independently reviewed positive net live performance, a limited premium cohort may receive the same evidence discipline and risk-controlled workflow. Until then, premium billing remains closed.