This page describes how content is produced, fact-checked, and updated on AI Code Invest. It exists to give readers a clear picture of editorial process and to be transparent about the role of AI tools in our workflow.
Editorial principles
Three principles guide every post on this site:
- Accuracy first. If we can’t verify it, we don’t publish it. Code that runs. Math that checks. Sources that exist.
- Practical over theoretical. Posts are written for someone who needs to do something with the information—implement an algorithm, evaluate a model, build a portfolio—not just understand it abstractly.
- Honest about uncertainty. When something is debated, we say so. When data is limited, we say so. When we don’t know, we say “we don’t know.”
How posts are produced
The blog is written by kongastral, a single human author. The end-to-end workflow for a typical technical post:
- Topic selection. Based on real questions encountered while building or investing, not chasen for SEO traffic alone.
- Source research. Original papers, official documentation, recognized textbooks, primary data. Reddit/Stack Overflow are not primary sources.
- Outline. A human outline that determines what the article will cover, what code examples are needed, and what figures will illustrate the concept.
- Drafting. AI tools (primarily Claude) assist with drafting prose from the outline, generating initial code snippets, and producing SVG diagrams. The author reviews and edits everything.
- Code verification. Every Python/SQL snippet that appears in a post is run locally. If it doesn’t work, it doesn’t ship.
- Math verification. Formulas are cross-checked against original sources.
- Source verification. External links are followed to confirm they exist and point to legitimate references.
- Final edit. Human read-through for tone, accuracy, completeness, and removal of generic AI patterns.
- Publish. Once it meets the bar, it goes live with the author byline.
Use of AI tools
We believe in being upfront about this. AI is part of the writing toolkit on this site, similar to how spellcheckers, grammar tools, and IDEs are part of a writer’s toolkit. Specifically:
- Drafting: AI helps turn outlines into first-draft prose. The author edits, restructures, and rewrites as needed.
- Code generation: AI helps produce starter code, which is then tested, debugged, and adapted by the author.
- Diagram generation: AI helps produce SVG figures from descriptions. The author validates that the figures correctly illustrate the concept.
- Research summarization: AI helps summarize long technical papers. The author then reads the actual paper for anything cited in a post.
AI is not used to:
- Invent statistics or quotes.
- Fabricate sources or references.
- Make claims that the author hasn’t independently verified.
- Replace editorial judgment about what to publish.
Fact-checking and corrections
Mistakes will happen. When they do:
- If you spot an error, please email kongastral@gmail.com or comment on the post.
- Substantive corrections (factual errors, broken code, mislabeled diagrams) are made within 7 days of being reported and noted with an updated date on the post.
- Minor edits (typos, phrasing) are made silently.
- Major retractions, if ever needed, will include a clear note explaining what changed and why.
Update cadence
Content ages. We try to keep posts current by:
- Updating posts when underlying technology changes (e.g., a new major library version, a new significant paper that changes best practice).
- Re-running code examples periodically to verify they still work with current package versions.
- Adding “Last updated” dates to posts that have been materially revised.
The “Last updated” date you see on a post reflects substantive content changes, not minor tweaks.
Sources and citations
Where a post relies on a specific paper, dataset, or external claim:
- The source is linked at the relevant point in the article and again in a References section at the end.
- Where possible, we link to the original paper or primary source—not aggregator sites.
- Where we paraphrase a source, we attribute it.
- We do not fabricate citations. If a claim isn’t backed by a real source, it isn’t published.
Conflict of interest disclosures
- The author does not receive payment for product placements or favorable coverage of any specific tool, broker, fund, or platform.
- When the author personally uses a tool mentioned in a post (e.g., a specific IDE or library), this is noted.
- The site shows Google AdSense advertising. AdSense placements are not aware of post content at editorial level—we do not write to chase ad rates.
- The author may hold individual U.S. stock positions. Investment posts are general educational content, not personalized recommendations, and posts about specific companies will note any author position when relevant.
Investment content notice
Comments and community
Constructive comments and corrections are welcome. We moderate to remove spam, abuse, and irrelevant promotional content. We do not delete comments simply because they disagree with a post’s conclusions.
Contact
For questions about content, corrections, or editorial policy, email kongastral@gmail.com.
Last updated: May 2026.