Everyday AI: Errata
Corrections to Everyday AI, Volume I: Basics of Artificial Intelligence, First Edition, by Dr. Nobel Khandaker.
A book about not trusting confident answers had better be willing to be corrected. This page lists confirmed errors as they are found, so that a reader who hits a passage that will not add up can check whether the problem is theirs or the book’s.
No errata reported yet for the First Edition.
That is a statement about what has been reported, not a claim that the book is perfect. If you have found something, the section below is how to tell us.
How to report an error
Please report anything that looks wrong. The small ones matter as much as the large ones, and because the book is still ahead of its first printing, a correction sent now can be fixed before it reaches paper.
The most useful reports are on GitHub Issues, because the thread stays visible to other readers who hit the same thing. If you would rather not use GitHub, email works too.
When you report, please include:
- Where. The chapter and section number — “Section 2.3” is unambiguous — and the printed page number if you have one.
- What it says. A short quotation is ideal.
- What you think it should say, and briefly why.
- Which format you were reading: the sample PDF, a page on this site, or the book itself.
What counts
Worth reporting:
- A factual claim that is wrong — a date, a figure, an attribution, a description of what a tool does.
- A dead or wrong citation. Chapter 1 alone carries several footnotes with live source links; a link that has rotted or that does not support the sentence it is attached to is an error worth fixing.
- A stale product fact. The book names real tools and quotes a real subscription price, current to mid-2026. When a price, a limit or a feature changes, that is a legitimate correction, and it is the kind this book will attract most.
- A definition that is technically wrong, or an analogy that misleads more than it explains.
- A cross-reference pointing at the wrong chapter, section or figure.
- Typos that change the meaning — a dropped “not” is worth more than a hundred transposed letters.
Probably not errata, though still worth raising:
- Style preferences, or a different way you would have explained something.
- Disagreement with a judgement call — which tool is better for which job, or how much to worry about a given risk. The book takes positions on purpose and says when it is doing so.
A note on shelf life
Some of what this book says was true when it was written and will stop being true — that is unavoidable in a book about a field moving this fast, and the book flags its own time-sensitive claims rather than pretending otherwise.
Corrections of that kind are still welcome here. A reader in 2028 hitting a 2026 price is better served by a note on this page than by silence.
Reported errors are credited by name on this page unless you ask otherwise.