husayn gokal
Geneva

← The Master Plan

Part 12 of 18

Research and Writing: Mini Essays, Literature, and Publishing

1. Purpose of This Part

This part defines the research and writing roadmap.

Research and writing are not separate from the rest of the master plan.

They are the system by which learning becomes contribution.

Software produces tools. EEE produces circuits. Physics produces models and simulations. Math produces proofs and methods. Cybersecurity produces findings and reports. Philosophy produces arguments and worldview change. Research and writing turn all of that into communicable knowledge.

The goal is:

To become someone who regularly turns curiosity, experiments, reading, building, failures, and insights into written intellectual output.

This connects directly to the original brief: the aim is to regularly write mini essays, develop ideas, possibly combine them into deeper work, and eventually publish research papers independently or collaboratively across the domains of the master plan.

The standard is:

Can I turn a question into sources, sources into understanding, understanding into experiments or arguments, and experiments or arguments into clear written contribution?

2. What Research Competence Actually Means

Research competence is not merely reading papers. It is not collecting PDFs.

It is not summarizing articles with AI.

It is not writing something that “sounds academic.”

Real research competence means being able to:

  • ask clear questions
  • search literature systematically
  • distinguish primary and secondary sources
  • read papers structurally
  • identify claims, methods, evidence, limitations, and gaps
  • track references properly
  • reproduce results where possible
  • document methods honestly
  • write clearly
  • cite accurately
  • disclose limitations
  • revise based on criticism
  • publish or share work responsibly

A serious researcher asks:

  • What exactly is the question?
  • Has someone already answered it?
  • What are the best sources?
  • What is the method?
  • What evidence supports the claim?
  • What assumptions are being made?
  • What is missing?
  • Can this be reproduced?
  • What would falsify or weaken the claim?
  • What contribution can I honestly make?
  • What should I not claim?

The standard is not:

    “Did I read a lot?”

The standard is:

    Can I produce a trustworthy, useful, well-sourced, honest piece of work?

3. The Research-Backed Source Spine

The research and writing roadmap should use scholarly search tools, reference managers, preprint repositories, open-science platforms, ethics guidance, and writing tools.

The main source spine is:

  • Google Scholar for broad scholarly search. Google Scholar describes itself as a way to

search scholarly literature across disciplines and sources, including articles, theses, books, abstracts, court opinions, academic publishers, repositories, universities, and professional societies. (Google Scholar)

  • arXiv for open-access preprints in fields directly relevant to this plan, including physics,

mathematics, computer science, statistics, electrical engineering and systems science, and economics. arXiv also states clearly that materials on the site are not peer-reviewed by arXiv, so papers must be read critically. (arXiv)

  • Semantic Scholar for AI-assisted scientific literature discovery. Semantic Scholar

describes itself as a free AI-powered research tool for scientific literature. (Semantic Scholar)

  • Zotero for reference management. Zotero describes itself as a free, easy-to-use tool to collect, organize, annotate, cite, and share research. (Zotero)
  • ORCID for researcher identity. ORCID provides a free, unique, persistent identifier for individuals engaged in research, scholarship, and innovation. (ORCID)
  • Zenodo for preserving and sharing research outputs. Zenodo describes itself as a

CERN service that helps researchers share and preserve research outputs in any size, format, and field, and Zenodo documentation explains that DOIs provide persistent links and improve discoverability. (GitHub)

  • OSF / Center for Open Science for open-science workflows. The Center for Open

Science states that its mission is to increase openness, integrity, and reproducibility in scholarly research. (Center for Open Science)

  • COPE for publication ethics. COPE says its role is to assist editors and publishers in preserving and promoting the integrity of the scholarly record. (publication-ethics.org)
  • ICMJE Recommendations for authorship, reporting, editing, publication ethics, and AI

use in publishing. The ICMJE recommendations were updated in January 2026 and explicitly include a section on AI use in publishing. (ICMJE)

  • EQUATOR Network for reporting guidelines, especially when work touches health,

psychology, experiments, interventions, or structured empirical reporting. EQUATOR describes itself as a resource for finding reporting guidelines and improving research writing. (EQUATOR Network)

  • Overleaf / LaTeX for technical writing and collaboration. Overleaf documentation

describes it as a collaborative LaTeX editor, with real-time collaboration, sharing, track changes, comments, and chat. (Overleaf Docs)

The rule is: Search broadly, read critically, cite carefully, write honestly, preserve outputs, and never fabricate scholarship.

4. The Researcher-Writer Identity

The identity to build here is:

     Independent research builder-writer.

This means you do not wait for an institution to give you permission to think seriously.

But it also means you do not pretend that independence removes standards.

Independent research must be more disciplined, not less.

A serious independent researcher-writer respects:

●​ sources ●​ evidence ●​ uncertainty ●​ authorship ●​ reproducibility ●​ citation ●​ peer criticism ●​ ethical limits ●​ clarity ●​ humility ●​ revision

This identity matters because the master plan spans many domains. Some outputs will be formal papers. Some will be technical reports. Some will be essays. Some will be simulations. Some will be literature maps. Some will be failed reproductions. Some will be case studies.

All of them should train the same habit:

Write so that another serious person can understand, verify, challenge, and build on the work.

5. The Research and Writing Roadmap

Ladder The roadmap is divided into layers.

Each layer must produce artifacts.

Do not move forward because you “researched a topic.”

Move forward when the research trail is visible.

Layer 0 — The Daily / Weekly Mini Essay

Habit Purpose The mini essay is the seed of the entire research system.

A mini essay is a short, focused piece of writing about one idea.

It should ideally fit on one screen.

It should take the pressure off “writing a paper” and instead build the habit of thinking in public or semi-public form.

Structure Each mini essay should include:

  1. Title
  2. One question
  3. Core idea
  4. Short explanation
  5. One example
  6. One implication
  7. One source or reference if relevant
  8. One unresolved question Example Topics
  • Why AI should not replace struggle in learning
  • What “building” means in philosophy
  • Why unit checks matter in physics
  • Why GitHub is a public memory system
  • What makes a cybersecurity report useful
  • Why RAG systems fail when retrieval fails
  • What a derivative means physically
  • Why op-amp idealizations are dangerous
  • Why quantum computing is not just faster computing
  • What counts as understanding a paper

Required Artifacts Create:

  1. Mini essay template
  2. Mini essay archive
  3. Weekly mini essay tracker
  4. Tag system by domain
  5. “Essay seeds for future papers” list
  6. Revision log
  7. Best essays folder
  8. Public/private decision system
  9. Essay-to-paper pipeline notes
  10. “Why I write” essay

Completion Standard This layer is complete when:

  • writing short essays becomes normal
  • each essay contains one clear idea
  • ideas are tagged and searchable
  • some essays naturally become larger projects
  • writing is no longer reserved for “when I know enough”

Layer 1 — Research Notes and Source

Discipline Purpose Research starts with disciplined note-taking.

Without source discipline, writing becomes unreliable.

The goal is to create notes that preserve where ideas came from, what the source actually says, and what your own interpretation is.

Note Types Use different note types:

  • source note
  • concept note
  • claim note
  • method note
  • quote note
  • objection note
  • experiment note
  • question note
  • synthesis note
  • personal reflection note

Source Note Template Each source note should include:

  1. Full citation
  2. Link or DOI
  3. Source type
  4. Field
  5. Main question
  6. Main claim
  7. Method
  8. Evidence
  9. Key terms
  10. Important figures/tables
  11. Limitations
  12. Your understanding
  13. Your doubts
  14. Related sources
  15. Possible use in future writing

Required Artifacts Create:

  1. Research note template
  2. Zotero library structure
  3. Tagging system
  4. Source reliability checklist
  5. Citation style notes
  6. Quote/paraphrase distinction guide
  7. Source-to-essay workflow
  8. Notes quality checklist
  9. Research notebook folder structure
  10. “How I avoid fake research” essay

Completion Standard This layer is complete when:

  • every important source has a traceable note
  • citations are stored properly
  • quotes and paraphrases are distinguished
  • personal ideas are separated from source claims
  • notes can be reused in essays and reports

Layer 2 — Literature Search

Purpose Literature search is how questions become situated in existing knowledge.

The goal is not to find one paper and stop. The goal is to map what already exists.

Tools Use:

  • Google Scholar for broad search across disciplines. (Google Scholar)
  • arXiv for preprints in physics, math, computer science, statistics, electrical engineering,

and related domains, while remembering that arXiv papers are not peer-reviewed by arXiv. (arXiv)

  • Semantic Scholar for AI-assisted discovery and citation exploration. (Semantic Scholar)
  • PhilPapers for philosophy literature.
  • official documentation and standards for technical domains.
  • references inside strong papers.

Search Method For each research question:

  1. Write the question plainly.
  2. Identify keywords.
  3. Identify synonyms.
  4. Search broad.
  5. Search narrow.
  6. Find review papers.
  7. Find recent papers.
  8. Find foundational papers.
  9. Follow citations backward.
  10. Follow citations forward.
  11. Save sources in Zotero.
  12. Create a literature map.

Required Artifacts Create:

  1. Literature search log
  2. Keyword/synonym table
  3. Foundational paper list
  4. Recent paper list
  5. Review paper list
  6. Citation chain map
  7. “What everyone agrees on” note
  8. “What is disputed” note
  9. “What is missing” note
  10. Literature search postmortem

Completion Standard This layer is complete when:

  • literature search is systematic
  • old and new sources are distinguished
  • foundational and recent works are both identified
  • citation trails are followed
  • the topic is mapped before claims are made

Layer 3 — Annotated Bibliographies and

Literature Maps Purpose An annotated bibliography turns a pile of sources into an organized research foundation.

A literature map shows how sources relate to each other.

Annotated Bibliography Entry Each entry should include:

  1. Citation
  2. Field/subfield
  3. Research question
  4. Main claim
  5. Method
  6. Evidence
  7. Strengths
  8. Weaknesses
  9. Relevance to your question
  10. How it connects to other sources Literature Map Categories Organize sources by:
  • theory
  • method
  • dataset
  • experiment
  • critique
  • application
  • historical foundation
  • recent development
  • unresolved problem
  • opposing position

Required Artifacts Create:

  1. Annotated bibliography template
  2. Annotated bibliography for AI topic
  3. Annotated bibliography for physics/quantum topic
  4. Annotated bibliography for EEE topic
  5. Annotated bibliography for philosophy topic
  6. Literature map diagram
  7. Research gap list
  8. Competing schools/approaches table
  9. Source quality ranking
  10. “What the literature seems to say” synthesis

Completion Standard This layer is complete when:

  • sources are summarized accurately
  • relationships between sources are visible
  • disagreements are identified
  • gaps are noted honestly
  • writing can proceed from a map, not chaos

Layer 4 — Reading Papers Structurally

Purpose A paper should not be read like a novel.

It should be dismantled.

The goal is to understand the structure of the contribution.

Universal Paper Reading Template For every paper:

  1. Citation
  2. Field
  3. Problem
  4. Why the problem matters
  5. Prior work
  6. Research gap
  7. Main claim
  8. Method
  9. Data or system
  10. Experiment or argument
  11. Metrics or evaluation criteria
  12. Results
  13. Figures/tables explained
  14. Key equations explained
  15. Limitations
  16. Threats to validity
  17. What I understood
  18. What I did not understand
  19. Possible reproduction
  20. Possible extension

Reading Passes Use three passes.

Pass 1 — Orientation

Read:

  • title
  • abstract
  • introduction
  • conclusion
  • figures
  • section headings

Goal:

    Know whether the paper matters.

Pass 2 — Structure

Read:

  • related work
  • method
  • experiments
  • results
  • limitations

Goal:

    Know what the paper claims and how it supports it.

Pass 3 — Deep Work

Study:

  • equations
  • algorithms
  • experimental setup
  • assumptions
  • code/data if available
  • reproducibility

Goal:

    Know whether you can explain, reproduce, or challenge the work.

Required Artifacts Create:

  1. Paper reading template
  2. Paper reading log
  3. Equation breakdown notebook
  4. Figure explanation notebook
  5. Method reconstruction notebook
  6. “What I did not understand” archive
  7. Reproduction candidate list
  8. Paper summary one-pagers
  9. Paper critique notes
  10. “How to read papers without pretending” essay

Completion Standard This layer is complete when:

  • papers can be summarized structurally
  • methods can be explained
  • figures are interpreted
  • equations are not ignored blindly
  • limitations are identified
  • possible reproductions emerge

Layer 5 — Reproducibility and Experiment

Logs Purpose Research becomes serious when claims can be tested, reproduced, or at least documented transparently.

The Center for Open Science explicitly focuses on openness, integrity, and reproducibility, which makes this layer central to trustworthy independent research. (Center for Open Science)

What Counts as Reproduction? Reproduction can mean:

  • rerunning the authors’ code
  • reproducing a figure
  • implementing a method from scratch
  • testing a claim on a small dataset
  • repeating a simulation
  • checking whether results are sensitive to parameters
  • reproducing an argument structure in philosophy
  • recreating a circuit behavior from a datasheet or paper
  • rebuilding an AI pipeline

Experiment Log Template Each experiment should include:

  1. Question
  2. Hypothesis
  3. Source or inspiration
  4. Materials/data/code
  5. Environment
  6. Method
  7. Parameters
  8. Results
  9. Plots/tables
  10. Failures
  11. Interpretation
  12. Limitations
  13. Next experiment
  14. Reproducibility notes

Required Artifacts Create:

  1. Experiment log template
  2. Reproduction attempt tracker
  3. Failed reproduction archive
  4. Dataset notes
  5. Code environment file
  6. Results folder structure
  7. Parameter log
  8. Figure reproduction notebook
  9. Reproducibility checklist
  10. “Failure as research output” essay

Completion Standard This layer is complete when:

  • experiments are logged clearly
  • failures are preserved
  • environments are recorded
  • code and data are organized
  • results can be rerun or audited
  • claims are not made beyond evidence

Layer 6 — Technical Reports

Purpose A technical report is the bridge between private notes and formal papers.

It is more structured than a blog post but less intimidating than a journal submission.

Technical reports are perfect for this master plan because many outputs may not immediately be publishable papers, but they still deserve serious documentation.

Technical Report Structure Use:

  1. Title
  2. Abstract
  3. Motivation
  4. Background
  5. Research question
  6. Method
  7. Implementation or experiment
  8. Results
  9. Discussion
  10. Limitations
  11. Future work
  12. References
  13. Appendices

Good Report Topics

  • Comparing RAG chunking strategies
  • Reproducing a LoRA experiment
  • Building and testing a local Ollama RAG assistant
  • Evaluating an AI study assistant
  • Simulating a physical oscillator
  • Measuring RC circuit behavior against simulation
  • Comparing PID tuning methods
  • Reviewing quantum hardware modalities
  • Mapping philosophical positions on AI understanding
  • Writing an OS shell implementation report

Required Artifacts Create:

  1. Technical report template
  2. AI technical report
  3. Physics simulation report
  4. EEE lab report
  5. Cybersecurity lab report
  6. Software architecture report
  7. Philosophy argument report
  8. Research limitation checklist
  9. Report peer-review checklist
  10. Technical report archive

Completion Standard This layer is complete when:

  • reports have clear structure
  • methods and limitations are explicit
  • claims are supported
  • figures and tables are explained
  • references are accurate
  • reports could be read by another serious person

Layer 7 — Review Papers and Survey

Essays Purpose A review paper synthesizes existing work.

A survey essay maps a field, problem, or debate.

This is useful when entering large domains such as quantum hardware, AI agents, semiconductor fabrication, philosophy of science, RAG evaluation, or cybersecurity methodology.

Review Structure Use:

  1. Topic
  2. Scope
  3. Search method
  4. Inclusion/exclusion criteria
  5. Thematic categories
  6. Major findings
  7. Disagreements
  8. Gaps
  9. Limitations of the review
  10. Future research questions

Possible Review Topics

  • Survey of RAG evaluation methods
  • Review of agent safety failure modes
  • Review of LoRA and PEFT methods
  • Survey of quantum computing learning resources
  • Review of superconducting vs trapped-ion qubits
  • Survey of semiconductor fabrication learning paths
  • Review of bug bounty reporting practices
  • Survey of philosophy of AI understanding
  • Review of open-source contribution models
  • Survey of math prerequisites for quantum computing

Required Artifacts Create:

  1. Review paper template
  2. Search protocol template
  3. Inclusion/exclusion table
  4. Literature matrix
  5. Thematic synthesis notes
  6. Gap analysis document
  7. Review paper draft
  8. Reference library
  9. Review limitations statement
  10. “What makes a review trustworthy?” essay

Completion Standard This layer is complete when:

  • the search process is documented
  • the scope is clear
  • sources are compared, not merely listed
  • gaps are identified
  • the review does not pretend to be exhaustive unless it actually is

Layer 8 — Original Research Projects

Purpose Original research begins when a question, method, experiment, or argument produces something genuinely new, even if small.

Original does not always mean revolutionary.

It may mean:

  • a new comparison
  • a new reproduction
  • a failed reproduction
  • a new dataset
  • a new tool
  • a new measurement
  • a new educational method
  • a new philosophical argument
  • a new synthesis
  • a new case study
  • a new benchmark
  • a new analysis of an existing method

Original Research Pipeline

  1. Start with mini essays.
  2. Identify repeated questions.
  3. Search literature.
  4. Build a literature map.
  5. Find a gap or unresolved issue.
  6. Design a small method or experiment.
  7. Run it.
  8. Document honestly.
  9. Write a technical report.
  10. Revise into a paper if strong enough.

Required Artifacts Create:

  1. Research question bank
  2. Candidate project tracker
  3. Feasibility assessment template
  4. Research proposal one-pager
  5. Method design notes
  6. Data/code plan
  7. Experiment plan
  8. Ethics/risk review
  9. Draft technical report
  10. Paper-readiness assessment

Completion Standard This layer is complete when:

  • research questions emerge naturally
  • literature is checked before novelty is claimed
  • methods are documented
  • limitations are honest
  • small contributions are valued
  • original work is not exaggerated

Layer 9 — Preprints, Publishing, and

Scholarly Identity Purpose Publishing is the act of making work available for others to read, evaluate, cite, challenge, and build on.

This does not require waiting for perfect institutional conditions.

But publishing must be done responsibly.

Scholarly Identity Create and maintain:

  • ORCID profile
  • serious GitHub profile
  • research writing archive
  • Zotero library
  • paper/project index
  • personal website research page
  • Zenodo archive for selected outputs
  • preprint plan where appropriate

ORCID is useful because it provides a persistent researcher identifier that connects individuals with their contributions across research, scholarship, and innovation. (ORCID)

Zenodo is useful for preserving research outputs, and its DOI documentation explains that DOIs provide permanent links, support citation attribution, interlink research outputs, and improve discoverability. (Zenodo)

Publishing Forms Outputs can include:

  • blog essays
  • GitHub reports
  • technical reports
  • whitepapers
  • preprints
  • workshop papers
  • conference submissions
  • journal submissions
  • datasets
  • code releases
  • replication reports
  • educational resources

Required Artifacts Create:

  1. ORCID profile
  2. Research page on personal website
  3. Publication tracker
  4. Preprint checklist
  5. Journal/conference fit checklist
  6. Zenodo/GitHub release workflow
  7. DOI and citation notes
  8. Author bio
  9. Research statement
  10. “Why publish?” essay

Completion Standard This layer is complete when:

  • serious outputs are findable
  • research identity is organized
  • selected work can be cited
  • publication decisions are intentional
  • public claims match evidence

Layer 10 — Research Ethics, Authorship,

and Integrity Purpose Research without integrity becomes noise or harm.

Ethics is not optional.

COPE focuses on preserving and promoting the integrity of the scholarly record, and ICMJE’s recommendations cover conduct, reporting, editing, publication, authorship, and AI use in scholarly publishing. (publication-ethics.org)

Ethics Topics

  • plagiarism
  • citation accuracy
  • authorship
  • contributorship
  • conflicts of interest
  • data fabrication
  • data falsification
  • image manipulation
  • p-hacking
  • selective reporting
  • duplicate publication
  • peer review ethics
  • AI disclosure
  • human subjects concerns
  • privacy
  • security-sensitive research
  • responsible disclosure
  • reproducibility

ICMJE recommends that authorship be based on four criteria and emphasizes that authorship implies responsibility and accountability for published work. (ICMJE)

Personal Research Integrity Rules

  1. Do not fabricate data.
  2. Do not invent sources.
  3. Do not cite papers not read or at least inspected responsibly.
  4. Do not hide negative results.
  5. Do not claim novelty without literature search.
  6. Do not use AI without verification.
  7. Do not list authors who did not contribute.
  8. Do not omit contributors unfairly.
  9. Do not overclaim.
  10. Do not publish security-sensitive details irresponsibly.

Required Artifacts Create:

  1. Personal research integrity policy
  2. Authorship policy
  3. AI-use disclosure policy
  4. Conflict-of-interest template
  5. Data handling policy
  6. Citation verification checklist
  7. Plagiarism avoidance guide
  8. Negative results policy
  9. Responsible publication checklist
  10. “Integrity over output” essay

Completion Standard This layer is complete when:

  • ethical rules are written
  • citations are verified
  • AI use is disclosed where appropriate
  • authorship is treated seriously
  • limitations and negative results are preserved
  • integrity matters more than looking impressive

6. Research and Writing Project Ladder

Research and writing should produce increasingly serious outputs.

Level 1 — Mini Essays Purpose: build writing rhythm. Examples:

  • one idea
  • one argument
  • one technical insight
  • one philosophical reflection
  • one experiment lesson

Level 2 — Source Notes Purpose: build evidence discipline.

Examples:

  • paper notes
  • textbook chapter notes
  • documentation notes
  • philosophy article notes
  • datasheet notes
  • standards notes

Level 3 — Literature Maps Purpose: understand fields.

Examples:

  • AI agents literature map
  • RAG evaluation literature map
  • quantum hardware literature map
  • semiconductor fabrication map
  • philosophy of science map
  • cybersecurity reporting map

Level 4 — Technical Reports Purpose: document projects seriously. Examples:

  • software architecture report
  • AI system evaluation report
  • physics simulation report
  • EEE circuit measurement report
  • cybersecurity lab report
  • OS project report

Level 5 — Reproduction Studies Purpose: verify and learn deeply.

Examples:

  • reproduce a paper figure
  • reproduce an ML experiment
  • reproduce a quantum circuit result
  • reproduce a circuit simulation
  • reproduce a statistical analysis
  • reproduce an algorithm benchmark

Level 6 — Review Papers Purpose: synthesize existing work.

Examples:

  • RAG evaluation review
  • AI agent safety review
  • quantum computing learning path review
  • LoRA/PEFT review
  • semiconductor fabrication overview
  • philosophy of AI review

Level 7 — Original Research Purpose: contribute new work. Examples:

  • new benchmark
  • new tool
  • new dataset
  • new comparison
  • new reproduction finding
  • new method variation
  • new philosophical argument
  • new educational framework
  • new case study

Level 8 — Published and Preserved Outputs Purpose: make work useful beyond yourself.

Examples:

  • GitHub release
  • Zenodo archive
  • preprint
  • conference submission
  • journal submission
  • personal website research page
  • open-source research tool
  • public technical report

7. Research GitHub and Public Output

Strategy Research output should be visible and organized.

Repository categories:

  1. mini-essays
  2. research-notes
  3. paper-reading-log
  4. literature-maps
  5. technical-reports
  6. paper-reproductions
  7. ai-evals-research
  8. physics-simulations-research
  9. eee-lab-reports
  10. philosophy-essays
  11. cybersecurity-reports-lab-only
  12. research-tools
  13. preprint-drafts
  14. publication-tracker

Each serious research repo should include:

  • README
  • research question
  • sources
  • method
  • data/code if applicable
  • environment setup
  • results
  • limitations
  • references
  • license
  • citation file where appropriate
  • reproducibility notes

The public-output goal is:

Make the research process visible enough that another person can inspect, learn from, challenge, or reuse it.

8. How Research Connects to the Other

Domains Software Development Research produces: - architecture reports

●​ system comparisons ●​ performance benchmarks ●​ open-source documentation ●​ product case studies

AI Research produces:

●​ evals ●​ model comparisons ●​ RAG experiments ●​ fine-tuning reports ●​ paper reproductions ●​ responsible AI analyses

Mathematics Research produces:

●​ derivation notes ●​ proof writeups ●​ numerical experiments ●​ optimization studies ●​ math-for-domain explanations

Physics Research produces:

●​ simulations ●​ experiment logs ●​ paper breakdowns ●​ quantum computing notebooks ●​ quantum hardware literature maps

EEE Research produces:

 - lab reports
  • PCB case studies
  • datasheet analyses
  • circuit comparisons
  • semiconductor fabrication notes

Cybersecurity Research produces:

  • authorized reports
  • defensive guidance
  • vulnerability class studies
  • methodology documents
  • responsible disclosure case studies

Philosophy Research produces:

  • argument maps
  • concept analyses
  • literature reviews
  • position papers
  • worldview revision logs

9. How AI Should Be Used in Research

and Writing AI can be extremely useful in research and writing, but it is also extremely dangerous.

It can fabricate sources.

It can flatten nuance.

It can summarize papers incorrectly.

It can make weak writing sound confident.

It can produce fake academic style. Therefore, AI must be used as a research assistant, not as the researcher.

Correct AI Use Use AI to:

  • brainstorm search terms
  • explain difficult passages
  • create reading questions
  • help structure notes
  • generate objections
  • compare paper methods
  • draft outlines
  • critique arguments
  • identify unclear writing
  • suggest missing limitations
  • generate reproducibility checklists
  • help format reports
  • help prepare peer-review questions

Incorrect AI Use Do not use AI to:

  • invent citations
  • summarize papers you did not inspect
  • fabricate data
  • fabricate experiments
  • fabricate peer review
  • hide uncertainty
  • write conclusions before results
  • claim novelty without literature search
  • replace source reading
  • produce final scholarship without verification

The AI Research Rule

AI may assist the research process, but every source, claim, citation, result, and conclusion must be human-verified.

For every AI-assisted research output:

  1. Save the prompt or assistance note if relevant.
  2. Verify all factual claims.
  3. Verify citations manually.
  4. Read or inspect original sources.
  5. Mark uncertainty.
  6. Disclose AI use where required.
  7. Rewrite in your own understanding.
  8. Do not publish unsupported claims.

10. Common Research and Writing Traps

Trap 1 — PDF Hoarding Collecting papers is not research.

Rule:

    Every saved paper needs a reason.

Trap 2 — Reading Without Notes Unrecorded reading disappears.

Rule:

    Every serious source gets a structured note.

Trap 3 — Citation Theater Citations should support claims, not decorate paragraphs.

Rule:

    Cite what actually supports the sentence.

Trap 4 — Overclaiming Small experiments do not justify massive conclusions.

Rule:

    Claims must match evidence.

Trap 5 — Fear of Publishing Waiting for perfection prevents growth.

Rule:

Publish appropriate outputs at appropriate levels: notes, reports, essays, preprints, or papers.

Trap 6 — AI-Written Scholarship AI can make nonsense sound polished.

Rule:

    No source, citation, or claim enters final writing without verification.

Trap 7 — No Reproducibility Untracked experiments cannot be trusted.

Rule:

    Code, data, parameters, environment, and failures must be logged.

Trap 8 — No Revision First drafts are not final thought.

Rule:

    Revision is where thinking becomes serious.

11. First 25 Serious Research and Writing

Artifacts These are the first serious artifacts for this domain.

Artifact 1 — Research and Writing Operating Manual A personal manual defining research workflow, writing workflow, citations, ethics, AI rules, and publication rules.

Artifact 2 — Mini Essay Archive A regular archive of short essays across all master plan domains.

Artifact 3 — Zotero Research Library A structured reference library with collections for AI, software, physics, EEE, cybersecurity, OS, math, philosophy, and research methods.

Artifact 4 — Source Note Template System Reusable templates for papers, books, documentation, standards, datasets, and philosophical texts.

Artifact 5 — Literature Search Log A record of search terms, databases, found sources, excluded sources, and search reflections.

Artifact 6 — Annotated Bibliography Pack Annotated bibliographies for at least five major domains.

Artifact 7 — Literature Map Repository Visual and written maps of research areas, debates, methods, and gaps.

Artifact 8 — Paper Reading Log Structured paper notes with problem, method, results, limitations, and possible reproductions.

Artifact 9 — Equation and Figure Breakdown Notebook A notebook for unpacking equations, graphs, tables, diagrams, and experimental results.

Artifact 10 — Reproduction Attempt Archive Successful and failed attempts to reproduce papers, figures, experiments, circuits, or simulations.

Artifact 11 — Experiment Log System A reusable structure for documenting hypotheses, methods, parameters, results, failures, and next steps.

Artifact 12 — Technical Report Template A serious report template for software, AI, physics, EEE, cybersecurity, OS, and research projects.

Artifact 13 — First AI Technical Report A report on an AI system, eval, RAG experiment, or model comparison.

Artifact 14 — First Physics Simulation Report A report documenting a simulation, method, results, and limitations.

Artifact 15 — First EEE Lab Report A report comparing circuit theory, simulation, and measured hardware behavior.

Artifact 16 — First Cybersecurity Lab Report A legal lab-only report with scope, methodology, evidence, impact, and remediation.

Artifact 17 — First Philosophy Position Paper A serious argument paper on knowledge, ethics, AI, science, or meaning.

Artifact 18 — Review Paper Draft A survey or review of a focused topic across one domain.

Artifact 19 — Research Question Bank A living document of possible research questions, sorted by domain and feasibility.

Artifact 20 — Personal Research Integrity Policy A written policy covering citations, authorship, AI use, conflicts, data, and limitations.

Artifact 21 — ORCID and Research Identity Setup A persistent researcher identity, research page, author bio, and publication tracker.

Artifact 22 — Zenodo / DOI Workflow Notes A workflow for preserving selected outputs and making them citable where appropriate.

Artifact 23 — Preprint Draft Folder A folder for work that may become formal preprints or submissions.

Artifact 24 — Peer Review and Feedback Log A record of critiques received, revisions made, and lessons learned. Artifact 25 — Research Maturity Review A long-form reflection on how research changed learning, building, writing, and intellectual honesty.

12. When to Move Forward

Do not move forward because a folder contains many papers.

Move forward when research behavior is visible.

Move past mini essays when:

  • short essays are written regularly
  • each essay has one clear idea
  • some essays become larger research seeds

Move past source notes when:

  • sources are cited properly
  • notes distinguish source claims from your interpretation
  • Zotero or another reference system is organized

Move past literature search basics when:

  • search terms are documented
  • foundational and recent sources are identified
  • citation chains are followed
  • gaps and disputes are visible

Move past annotated bibliographies when:

  • sources are summarized accurately
  • relevance is explained
  • strengths and weaknesses are noted
  • sources are grouped thematically Move past paper reading basics when:
  • papers can be structurally analyzed
  • methods and limitations are understood
  • figures and equations are explained
  • possible reproductions are identified

Move past reproducibility basics when:

  • experiments have logs
  • code/data/environments are documented
  • failures are preserved
  • results can be rerun or inspected

Move past technical reports when:

  • reports have clear methods, results, limitations, and references
  • claims match evidence
  • another person could understand the work

Move into review papers when:

  • literature maps exist
  • the scope is clear
  • sources are compared and synthesized
  • gaps can be stated honestly

Move into original research when:

  • a question is grounded in literature
  • a method is feasible
  • contribution is small but real
  • limitations are understood

Move into formal publishing when:

  • the work is structured
  • sources are verified
  • claims are responsible
  • ethics and authorship are clear
  • feedback has been considered

13. The Research and Writing Standard

The final standard for this domain is:

I can ask serious questions, search literature systematically, read sources critically, take disciplined notes, build literature maps, run or reproduce experiments, write technical reports, produce essays, preserve research outputs, publish responsibly, and contribute honestly to knowledge.

Research is not separate from building.

It is how building becomes knowledge.

Writing is not separate from thinking.

It is how thinking becomes inspectable.

Publication is not vanity.

It is how useful work becomes available to others.

The long-term result should be a life where curiosity does not disappear into private excitement.

It becomes notes. Notes become essays. Essays become experiments. Experiments become reports. Reports become papers. Papers become tools, questions, collaborations, and contributions.