Get AI summaries of any video or article — Sign up free
This unbelievable AI tool makes publishing papers easy thumbnail

This unbelievable AI tool makes publishing papers easy

Academic English Now·
5 min read

Based on Academic English Now's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Use AI as an editor and brainstorming partner, not as a ghostwriter for full-paper submission.

Briefing

A practical workflow for writing research papers with AI—without plagiarism—centers on using AI as an editor and idea partner, not as a ghostwriter. The core message is that anti-plagiarism and AI-detection systems are now strong enough that “letting AI write the whole paper” is both unethical and risky. Instead, AI should improve readability, help brainstorm and structure sections, and generate drafts or expansions that the researcher must then verify, rewrite in their own words, and support with real sources.

The transcript lays out six ethics-and-integrity rules drawn from major scientific publishing norms: use AI only to improve readability; never generate entire text for submission; treat AI as a 24/7 colleague for brainstorming and questions; verify every AI output; draw final conclusions from the underlying data; and create original figures, images, and tables yourself. These constraints set up the tool demonstration: Jenny is presented as the “ultimate” research-writing assistant because it can speed up the most time-consuming parts of academic writing while still requiring the user to do the intellectual work.

The first bottleneck is paper structure. Writing without a clear outline can lead to weeks of work collapsing into a rewrite. Jenny is used to generate a detailed outline for a specific section (and potentially the whole paper) by prompting it with the topic, field, aim, and target length. The transcript emphasizes that prompt detail matters: specifying the section type (e.g., literature review), the study’s goal, and approximate length helps Jenny produce more accurate headings and subheadings than generic templates. Jenny can also provide suggested references per section when the user has uploaded materials.

The second bottleneck is the loneliness of writing. When researchers get stuck defining terms, finding precise details, or locating where an idea came from, they often lose momentum by searching through papers. Jenny is positioned as an always-available assistant for quick definitions and clarifications—followed by a warning that outputs must be verified and rewritten. The same “speed without surrendering authorship” approach is applied to expanding ideas: selecting a short passage and using commands like “write with more depth” to turn two sentences into a fuller subsection, or generating opposing arguments to reflect lack of consensus in the literature review.

A third pain point is citation recall—knowing something is true but forgetting which paper supports it. Jenny can suggest references if the user imports a BibTeX file (from tools like Zotero or Mendeley) or uploads PDFs directly. The transcript recommends using the user’s own library for safer, more accurate sourcing, while still requiring the researcher to check relevance before citing.

Finally, Jenny is used for polishing: improving fluency, simplifying or summarizing overly long passages, and tightening language to match expectations of high-tier journals indexed by Scopus and Q1 rankings. The transcript closes with a caution that no AI tool guarantees acceptance—since many submissions are rejected—so a broader publishing system matters beyond drafting and editing.

Cornell Notes

The transcript argues that AI can help researchers write papers faster and more clearly without plagiarism when it’s used as an editor, brainstorm partner, and citation assistant—not as a ghostwriter. It lays out strict ethics rules: improve readability only, never submit AI-written full text, verify every AI output, draw conclusions from the researcher’s own understanding, and create original figures and tables. Jenny is demonstrated as a workflow tool: generating detailed outlines from prompts, answering research questions (with verification), expanding short sections into deeper paragraphs, producing opposing arguments for literature reviews, and suggesting references after importing PDFs or BibTeX. The payoff is reduced time spent on structure, definitions, and source-finding, plus final language polishing via fluency and simplification tools.

What are the key rules for using AI in academic writing without plagiarism or detection trouble?

The transcript lists six rules: (1) use AI only to improve readability (proofreading/editing), (2) don’t use AI to write entire text for submission, (3) use AI to brainstorm ideas and answer questions like a 24/7 colleague, (4) verify AI output instead of trusting it blindly, (5) draw your own conclusions from the data, and (6) create your own figures, images, and tables rather than generating them with AI.

How does Jenny help with the most common early-stage writing failure—getting stuck on structure?

Jenny is used to generate an outline from a detailed prompt. The prompt should include the text type, field of study, the aim of the paper, and the target length, plus any required elements (e.g., limiting to the literature review). Jenny then produces headings and subheadings that are more specific than generic section labels, and can estimate how long each section should be based on the total length.

What does “verify and rewrite” mean in practice when Jenny provides definitions or expanded text?

When Jenny gives a definition (example: clarifying “Native speakerism”) or expands a short passage, the researcher must check that the details match the underlying literature and then rewrite in their own words. The transcript warns that AI output can be wrong or become risky if treated as final text, so verification is required before using it in the paper.

How can Jenny support literature reviews beyond outlining—especially when there’s no consensus?

For literature reviews, Jenny can generate opposing arguments. The transcript frames this as useful because it signals that the writer has read widely and understands disagreements in the field. As with other outputs, those arguments must be verified and then integrated through the researcher’s own wording and reasoning.

How does Jenny help with citations when the writer can’t remember which paper supports a claim?

Jenny can suggest references if the user uploads PDFs or imports a BibTeX file (the transcript mentions exporting from Zotero or Mendeley). In “library mode,” Jenny draws from the user’s uploaded documents, which the transcript calls safer and more accurate. In “discover mode,” Jenny proposes additional relevant papers, but the user still must verify that the suggested sources truly match the sentence or paragraph being cited.

What final editing steps does the transcript recommend after drafting with Jenny?

After drafting, Jenny can polish language using commands such as “improve fluency” (fixing grammar, awkward phrasing, and sentence connections) and “simplify” or “summarize” to reduce verbosity. The transcript emphasizes that high-tier journals value concise, clear expression of complex ideas, so tightening long sections can improve readability and alignment with journal expectations.

Review Questions

  1. Which of the six AI-use rules most directly prevents plagiarism, and why?
  2. How would you design a prompt for Jenny to generate a literature review outline that matches a specific word count?
  3. What steps should be taken before citing references suggested by Jenny in library mode versus discover mode?

Key Points

  1. 1

    Use AI as an editor and brainstorming partner, not as a ghostwriter for full-paper submission.

  2. 2

    Follow the six integrity rules: improve readability only, never generate entire text, verify output, draw your own conclusions, and create original figures/tables.

  3. 3

    Generate a detailed outline first by prompting Jenny with field, aim, and target length to avoid weeks of rewriting.

  4. 4

    Use Jenny to answer definition and detail questions quickly, but always verify and rewrite in your own words.

  5. 5

    Expand weak sections by selecting text and using depth commands, and strengthen literature reviews by generating opposing arguments (then verifying).

  6. 6

    Import your own PDFs or BibTeX so Jenny can suggest references; still check relevance before citing.

  7. 7

    Polish for journal readability with fluency improvements and simplification/summarization to reduce unnecessary “waffling.”

Highlights

AI use is framed as an ethics problem: anti-plagiarism and AI-detection tools are strong enough that full AI-written submissions are both risky and dishonest.
Jenny’s outline workflow depends heavily on prompt specificity—field, aim, and length drive better headings than generic templates.
Citation recovery is treated as a workflow: upload PDFs or import BibTeX so Jenny can suggest sources, but verification remains mandatory.
The transcript positions opposing-argument generation as a literature-review strength, reflecting real disagreement rather than one-sided claims.

Topics

Mentioned