How to Compare AI Stock Forecasts: CAGR and Total Return

Why annualized gains, five-year gains and averaged opinions need different labels, with an original Batch 7 example.

Oct 2, 20269 min readiPulse AI

Concepts library

Build durable understanding

6 articles. Fundamental ideas, classifications and foundational experiments in AI and forecasting that underpin formal research.

Methodology & Explainability

Why annualized gains, five-year gains and averaged opinions need different labels, with an original Batch 7 example.

Oct 2, 20269 min readiPulse AI

Product, Decisions & Building

A practical guide to choosing agent architectures by task shape, managing coordination costs, and keeping one accountable release path.

Sep 19, 202612 min readTowards AI

Consensus & Disagreement

Across 377 assets, a simple vote and a risk-aware scoring system disagreed in 45% of decisive cases. An audit for builders of agent teams.

Sep 19, 202612 min readLevel Up Coding

Consensus & Disagreement

What SpaceX, PetroChina, and NuScale reveal about using model disagreement as a research tool instead of treating consensus as truth.

Jul 20, 20267 min readDataDrivenInvestor

Methodology & Explainability

The problem with collapsing return, confidence, risk, financial strength, and catalysts into one number.

Jul 16, 20267 min readFinancial Strategy

Product, Decisions & Building

What three years of building iPulse AI taught us about trust, judgment, and the future of professional analysis

Jul 3, 202610 min readBootcamp

iPulse AI Investment Intelligence

Substack is the newsletter and discussion layer. This publication library remains the durable owned research record.