Call for PapersJournal of Measured AI Practice

What we publish — all four

1. Discovery. Something you did not know before the conversation with AI. Say what changed.
2. Discussion. You challenged, tested and revised the AI's answers — show the trace.
3. Data. Every claim rests on data or a source anyone can re-check.
4. Logic. Question → method → result → limitations, written honestly, including what failed.

What we do not publish

• Text copied from an AI answer
• Opinion or assertion without data
• Mere confirmation of a known fact
• Conclusions drawn only from synthetic examples

Open to everyone

• Age, degree, affiliation and nationality do not matter. Write in any language — AI reads and translates it, and your original is published alongside.
• Review is kind: if we cannot publish, we tell you concretely what would make it publishable.
• Authors under 14 need a guardian's consent (privacy). Pen names are allowed.

Transparency

Name the AI tools used and how; attach key conversation excerpts. Judgment and responsibility stay with the author.

Article types

• Research Article — full study with data and code.
• Practice Note — one tool or workflow, measured in real use (2–6 pages).
• Negative Result / Model Revision — a hypothesis that failed, and what replaced it.
• Registered Report — protocol published before data are seen; results follow.
• Correction — what we got wrong in an earlier piece.

Requirements

1. Real data. Conclusions drawn only from synthetic examples are not accepted.
2. Criteria before data. State how you decided success before looking (pre-registration encouraged; a hash or dated file is enough).
3. Receipts. Share data and code, or explain exactly why not. Remove personal and employer-confidential data.
4. AI disclosure. Say which AI tools did what (analysis, writing, translation).
5. Prior posting is fine. arXiv/SSRN preprints are welcome; state where. Do not submit work under review elsewhere if that venue forbids it.

Process

Submit in any language → editor desk check within 7 days → evidence check (every numeric claim traced to data or quote) + at least one reviewer → decision → published in English (default) with Korean and the author's original language side by side — the original is kept so anyone can check the translation.

Fees and rights

No submission or publication fees. Free to read. Authors keep copyright. (Open license: to be announced.)

Writing with AI — template & guide

HEARIM is a journal for people who write with AI. Your AI is the first reader of this template; humans are the last. So the template does two jobs:

• For your AI — a structured instruction block (role, rules, step-by-step order: outline → one section at a time → self-check). It forbids invented data and references and makes every number point back to your files.
• For your human readers — a style guide and a readability checklist: finding first, one idea per paragraph, terms explained, numbers with units, no AI filler phrases.

How to use: paste the “Instructions for the AI” block into your AI, attach the template with your data and conversation logs, and go section by section.

• Markdown (.md) — best for AI
• Word (.docx)
• 한국어 템플릿 (.md)

Sections: Abstract · Introduction (what you did not know) · Data & Method (criteria written before looking) · Results · Discussion (arguing with the AI) · Limitations & what failed · Conclusion · Statements · References · Appendix A (AI conversation excerpts).

Manuscript format

HTML is the version of record. Send one self-contained .html file with every figure embedded (base64). Quarto: embed-resources: true; Pandoc: --embed-resources --standalone. A PDF is welcome as a companion copy of the same text.

• HTML — at most 5 MB. Each figure at most 1600 px wide and 300 KB: JPEG/WebP for photos, PNG/SVG for charts.
• Keep it plain. No video, audio, iframes, trackers, extra embedded fonts or custom JavaScript. Standard math rendering is fine. We remove scripts and re-compress images before publishing.
• PDF — at most 5 MB. Searchable text (no scans), fonts embedded, images at most 300 dpi (vector charts preferred). Same content as the HTML.
• Too large? We send it back with a one-line fix, e.g. gs -sDEVICE=pdfwrite -dPDFSETTINGS=/ebook -o small.pdf in.pdf, or re-export figures at a smaller size.

Readable on phone, tablet and laptop

Most readers open HEARIM on a phone. We apply a house stylesheet (text size, table scrolling, tap-to-zoom figures), so do not add your own CSS. Make the figures themselves work at phone width:

• Shape. 4:3 or square works on every screen; avoid very wide panels (wider than 2:1). Split a 6-panel figure into two.
• Text inside figures. Export at the final size with labels of at least 11 pt; keep axis labels horizontal where possible; put the legend outside the data.
• Tables. At most 5 columns; move long explanations out of cells into notes below the table.
• Colour. Do not rely on colour alone (use line style or labels); check it in grayscale.

/* phone  */ @media (max-width:600px)                    { body text 17px, figures full width, tables scroll }
/* tablet */ @media (min-width:601px) and (max-width:1024px){ body text 17.5px }
/* laptop */ @media (min-width:1025px)                    { body text 17px, line length 72ch, figures ≤ 900px }

Submit

scholarbridge2026@gmail.com

Subject line: HEARIM submission — <title>

Include: manuscript (HTML of record + optional PDF, 5 MB each), data/code link or statement, AI-use statement, AI conversation excerpts (appendix), prior-posting statement.

• Optional: links to your social-media profiles (LinkedIn, X, GitHub…). We show them under your paper; free reader discussion happens there, tagged #HEARIM.