HOW THE RESEARCH WORKS
How AI Economy Radar researches AI signals
AI Economy Radar is an independent AI investment research publication. Its analyses link dated evidence to a specific business or market question so investors and researchers can test a thesis and see where uncertainty remains.
Start with the source and the date
Research articles link to the materials used for their claims, such as company filings and disclosures, official statistics, company or provider materials, and datasets where relevant. The source, reporting period and measure are identified alongside the observation so readers can check the original context.
A published value describes what its source reports for that period. It does not by itself establish a forecast, profitability, market share or investment return.
Cross-check evidence, not repetitions
Where the question calls for it, an analysis compares evidence from different parts of the AI economy—for example, adoption with provider revenue, demand with infrastructure delivery, or capital spending with its funding. Repeated views of the same underlying dataset are not treated as independent confirmation.
Differences in definitions, periods, sample coverage and reporting lags can limit a comparison. Articles describe those limits where they affect the conclusion.
Separate observation from inference
Reported facts are tied to their cited sources. Interpretation explains how those facts may bear on an investment thesis, while noting material caveats and alternative explanations. A research inference is not presented as a fact established by the source.
The analyses are evidence for further research, not personalized investment advice or a promise about future performance.
The tracker contains dated historical previews
Each available edition shows its snapshot date and data-coverage date, along with the definitions and sources for that view. The collection is not a live feed, and a past edition should not be read as a statement of current values.
Research articles display their publication dates. New articles and tracker editions appear when published; a historical edition is not silently refreshed or presented as current.
Social Radar uses captured public stock-call samples
Social Radar is a bounded collection of captured public posts and related stock-call records. Original-post and transcript coverage can be partial; the newest captured item is not guaranteed to be the newest item published. Missing or unreviewed history can create selection bias, so its records are not a complete two-year history or an unattended live recommendations feed.
Its sample metrics count verified directional stock recommendations with a finished return window. Sponsored content, cryptic posts, policy events and unverified attribution are excluded from those metrics. Raw adjusted stock-price change is shown separately from direction-adjusted call results and same-direction comparison with SPY. Entry and exit windows follow the displayed post timestamp and trading-session definitions.
Captured post performance does not establish that a post caused a price move, and the sample is not an audited strategy backtest.
Publication dates and corrections
Dates on research pages identify when an analysis was published. Source links remain with the analysis so readers can check underlying evidence. The site does not currently provide a separate public correction log; an article should be read with its displayed date and cited source context.