A restrained machine-intelligence command network with concentric control rings.

The Automated Auxiliary

The AI Tactical Ally

Deploying machine intelligence on the financial battlefield.

“If you know the enemy and know yourself, you need not fear the result of a hundred battles.”

AI can organize public information, accelerate screening, test explicit rules, and help standardize discipline. It does not know the future, uncover private institutional intent, or remove market risk.

Command principle: Machine intelligence serves the written plan. The human operator retains authority, verification duties, and accountability.

Foreknowledge Adaptation Discipline

The Strategic Premise

AI is an auxiliary force—not the general.

The advantage is not “machine certainty.” The advantage is a more disciplined workflow: faster triage, consistent rule application, documented assumptions, repeatable testing, and the ability to abstain when evidence is weak.

Sun Tzu Trading therefore evaluates every AI tool by four questions: What data does it use? What decision does it support? Where can it fail? Who remains accountable when it does?

The Three AI-Warfare Doctrines

Foreknowledge. Adaptation. Discipline.

Each doctrine converts an ancient strategic principle into a modern operating rule with explicit limits.

Doctrine 01

The Machine Scout

“What enables the wise sovereign and the good general to strike and conquer … is foreknowledge.”Sun Tzu, The Use of Spies

Doctrine

Use machine learning, structured data, and automated screening to narrow a large field into a smaller research queue.

Tactical application

A disciplined AI workflow can scan many instruments for user-defined conditions, compare liquidity and volatility, and surface anomalies for human review. It cannot reveal hidden intent or guarantee that an observed pattern will persist.

01Define the data universe before scanning02Require explainable screening criteria03Verify anomalies against primary market data04Reject outputs that cannot be independently reproduced

Simulate the Battle

Backtest the rule. Forward-test the process. Distrust the illusion.

Historical simulation can expose obvious weaknesses, estimate sensitivity to assumptions, and compare risk profiles. It cannot reproduce every fill, cost, regime change, or behavioral response that will occur in live markets.

  1. 01
    Define the hypothesis

    Specify the universe, signal, sizing, costs, exits, and invalidation criteria before examining results.

  2. 02
    Separate development from validation

    Reserve unseen data and test multiple regimes to reduce hindsight and overfitting.

  3. 03
    Stress the failure modes

    Model gaps, slippage, missing data, delayed signals, correlated losses, and execution outages.

  4. 04
    Forward-test without urgency

    Use simulation or minimal-risk observation before considering wider deployment.

Rules of Engagement

Six guardrails before any AI-assisted campaign.

01

No prophecy

AI outputs are estimates, classifications, summaries, or generated text—not foreknowledge of future prices.

02

No blind execution

Do not connect an unverified model or third-party service to capital without controls, monitoring, and kill switches.

03

No hidden assumptions

Document data sources, lookback periods, fees, slippage, constraints, and any human overrides.

04

No sensitive disclosure

Do not paste brokerage credentials, account statements, private keys, or confidential personal information into public AI systems.

05

No performance worship

Backtests and model scores can be optimized by hindsight and may fail when the regime changes.

06

Human command remains

The trader is responsible for verification, suitability, risk limits, compliance, and the decision to act or remain flat.

Candidate Resource Directory

Systems under review for the tactical stack.

These providers are included as research candidates, not endorsements. Commercial buttons remain inactive until independent review, formal partner approval, and exact offer verification are complete.

No active affiliate relationships in this directory.

Sun Tzu Trading does not currently rank, guarantee, or recommend these services. Features, pricing, eligibility, data coverage, and broker integrations can change. Verify all claims directly and evaluate suitability independently.

01

Resource Theater

AI Charting & Pattern Recognition

TrendSpiderCandidate

Automated chart analysis, market scanning, alerts, and strategy testing.

Best for
Traders who want to reduce repetitive chart work while retaining control of the thesis and risk plan.
Primary caution
Validate every alert and model output. Automated analysis can be delayed, incomplete, or wrong.
TickeronCandidate

Model-driven pattern search, market-screening tools, and AI-agent research workflows.

Best for
Comparing machine-generated pattern classifications with an independently prepared trade plan.
Primary caution
Do not rely on promotional performance claims. Review methodology, assumptions, costs, and live results independently.
02

Resource Theater

No-Code Automation & Simulation

ComposerCandidate

AI-assisted strategy construction, backtesting, and rules-based execution without traditional coding.

Best for
Systematizing portfolio rules and testing repeatable decision logic before considering live deployment.
Primary caution
Backtests are hypothetical and may reflect hindsight, overfitting, survivorship bias, or unrealistic execution assumptions.
Option AlphaCandidate

User-configured options and equity bots with no-code decision logic and risk controls.

Best for
Translating a documented options process into repeatable automations while preserving human oversight.
Primary caution
Automation can amplify a flawed rule. Broker connectivity, liquidity, fills, assignment, and platform availability remain material risks.
03

Resource Theater

Sentiment & News Intelligence

Provider Under ReviewCandidate

Potential monitoring of public news, filings, and crowd-sentiment shifts for research triage.

Best for
Prioritizing what deserves human investigation—not converting sentiment scores into automatic trade instructions.
Primary caution
Sentiment data can be noisy, manipulated, incomplete, stale, or misclassified. No provider has been approved for this category.

Sources & Verification

Primary texts, regulatory cautions, and provider documentation.

The philosophical quotations use Lionel Giles’s public-domain translation. Product capabilities are described conservatively from official provider materials; no performance claims are adopted.

Final Command

The strongest AI decision may be to do nothing.

A disciplined auxiliary should improve selection, verification, sizing, and restraint. When the evidence is weak, the data is compromised, or the risk cannot be bounded, the correct output is not a trade—it is abstention.