Updated October 5, 2026
Editorial Standards and Updates
Strategy Research Lab is independently operated. Its operator is responsible for publication choices, explanations and corrections. This page describes how readers can inspect the evidence and report a problem.
What a published study must explain
- The specific question, sample and time boundary.
- Rules, costs, data sources and what was knowable at the time.
- Results alongside counterevidence, limitations and failure conditions.
- Evidence a reader can inspect, and a way to report errors.
Historical strategy pages and downloads have been retired from the public site. The project repository retains research files for version checks. Shared NAV series, parameter variants and partial samples are not independent confirmations.
Sources and automation
Prefer issuer and regulatory disclosures or original research. Attribute secondary reporting and preserve uncertainty. Facts, interpretations and hypothetical scenarios must remain distinct. Images retain source and license information when reused.
AI tools assist with coding, research organization, translation and drafting. Automated calculations and structure checks do not constitute independent factual verification. Market Radar’s feed headlines and automated briefs are research aids; they do not establish the facts behind the linked stories.
Updates and corrections
Published and updated dates are shown with each article. Strategy evidence downloads state their frozen period; later tracking data carry their own cutoff. Material corrections should identify the affected claim, source and date rather than silently rewrite the research history.
For an error, send the page URL, disputed statement and verifiable source to [email protected], or use the feedback page. Avoid posting personal information publicly.
Site maintenance record
— Revised both device articles in Chinese and English around AI demand, storage prices and local AI upgrade decisions. Added dated Micron and TrendForce sources, separated company results and contract-price forecasts from retail prices, and checked local AI guidance against Ollama documentation. These articles contain no retail price tracking or hands-on benchmarks.
— Published Chinese and English articles on USB-C hub compatibility and portable SSD selection, with official specification sources and a clear distinction between document review and hands-on testing. These editions contain no affiliate product links.
— Edited all four essays in Chinese and English for clearer openings, shorter explanations and more natural phrasing. Updated article summaries and clarified the distinction between policy facts and interpretation. This was a prose revision: calculations and cited data were not refreshed, and original source cutoffs remain in place.
— Added dated, source-linked company research pages; moved the full headline ledger outside indexed pages; removed subjective profit percentages from Market Radar; and required a separate editor approval of the final bilingual draft before timed publication.
— Retired public strategy comparison, archive, trade and download URLs with an explicit 410 response; kept research files in the repository. The public site now directs readers to Market Radar and the blog.
— Moved strategy tools and the daily puzzle outside the site's main navigation and sitemap; retained legacy URLs with noindex and removed advertising from those pages. Added a local, content-addressed blog publication path that refuses drafts without source and rights records.
— Expanded the homepage with an evidence-first reading path and introductions to reviewed articles, so visitors can reach dated sources and research boundaries directly from the entry page.
— Separated archive charts from strategy selection, grouped shared NAV versions, narrowed publication and ad eligibility, exposed research evidence downloads, and clarified feed and author responsibilities.
Published articles and content dates
- An orbital TPU answered a prompt. What would establish sustained computing?
Published 2026-11-01 · Content updated 2026-11-01 - Which data should healthcare institutions make available to AI?
Published 2026-10-30 · Content updated 2026-10-30 - How can buyers assess local language AI?
Published 2026-10-27 · Content updated 2026-10-27 - How should businesses measure open-weight model savings?
Published 2026-10-26 · Content updated 2026-10-26 - What evidence would justify reducing physical tests with synthetic data and digital twins?
Published 2026-10-25 · Content updated 2026-10-25 - When AI groups alerts, how much should operators trust the diagnosis?
Published 2026-10-24 · Content updated 2026-10-24 - When do capacity advances become usable AI compute?
Published 2026-10-21 · Content updated 2026-10-22 - When is sharing telecom infrastructure with AI worth the investment?
Published 2026-10-22 · Content updated 2026-10-22 - What control should organizations retain when agents handle internal material?
Published 2026-10-20 · Content updated 2026-10-20 - What evidence would show that an extra driving safety stack adds protection?
Published 2026-10-19 · Content updated 2026-10-19 - Mercedes-Benz and Volvo plan to use similar driving technologies. What can buyers infer?
Published 2026-10-18 · Content updated 2026-10-19 - Is RTX Spark worth replacing a working creative PC?
Published 2026-10-17 · Content updated 2026-10-17 - Can small labs use Europe's expanding AI compute?
Published 2026-10-16 · Content updated 2026-10-16 - If selected etch modeling gains up to 30x as planned, how much could the factory gain?
Published 2026-10-15 · Content updated 2026-10-15 - How much do AI cloud growth figures tell us about investment returns?
Published 2026-10-14 · Content updated 2026-10-14 - When does Starship’s launch capacity become usable Starlink service?
Published 2026-10-13 · Content updated 2026-10-14 - What must Rocket Lab's orders deliver for shareholders to benefit?
Published 2026-10-12 · Content updated 2026-10-12 - Where US-China decoupling lands: policy documents, Chinese factories and firms
Published 2026-10-09 · Content updated 2026-10-10 - A-share reversal on October 9: green indexes, bruised tech
Published 2026-10-09 · Content updated 2026-10-09 - The Fed raised rates 25 basis points. Whose borrowing cost changed?
Published 2026-10-08 · Content updated 2026-10-08 - A 105GB model on a 32GB Mac: does a second SSD help?
Published 2026-10-07 · Content updated 2026-10-07 - Paying the bills with A-share dividends
Published 2026-10-06 · Content updated 2026-10-06 - The AI storage cycle: from four companies’ results to an SSD purchase
Published 2026-10-05 · Content updated 2026-10-05 - The local AI upgrade budget: capacity, bandwidth, context and task cost
Published 2026-10-05 · Content updated 2026-10-05 - Token is not a wage: from piecework to agents, what are firms actually buying?
Published 2026-09-19 · Content updated 2026-10-05 - Bank stocks: dividends or patience? A decade of data on A-share core holdings
Published 2026-09-18 · Content updated 2026-10-08 - When AI moves faster than society: who sets the pace, and who bears the cost?
Published 2026-09-15 · Content updated 2026-10-05 - How to read a strategy backtest: trace the conclusion back to the evidence
Published 2026-09-15 · Content updated 2026-10-08