2 min read

Two New GPTs

Two New GPTs

I have created two new GPTs for the store. Both experiments to see what we can do. If anyone wishes to go deeper into any of these ideas let me know.

Publish or Perish GPT

Publish or Perish is designed to help researchers format their manuscripts according to the specific guidelines of Nature and PLOS ONE journals. The app contains a comprehensive taxonomy of detailed formatting and submission requirements for these two academic journals which I created.

It is a prototype! If you're preparing your an article for submission to either of these journals, this GPT could be immensely helpful. Here’s what it does:

  1. Target Journal Confirmation: First, it asks you to confirm whether your manuscript is intended for Nature or PLOS ONE.
  2. Manuscript Upload Request: It then asks you to upload your manuscript.
  3. Guideline Reference: Robot Chef's Publish or Perish has a knowledge base of the detailed formatting and submission guidelines for both journals which I created. It uses this information to review your manuscript.
  4. Comprehensive Analysis: The GPT will analyze your manuscript, focusing on compliance with each specific guideline of the chosen journal. This includes checking the formatting of species names, gene nomenclature, manuscript organization, title specifications, authorship details, and more, depending on the journal.
  5. Checklist Report Generation: After the analysis, it generates a report in a checklist format, marking each guideline as compliant or non-compliant.
  6. Detailed Feedback: For each non-compliant item, the GPT provides detailed feedback, suggesting specific revisions or improvements.
  7. Focused Analysis: The analysis is strictly limited to the guideline compliance and does not include additional content review or overviews.

This GPT is particularly useful for ensuring that the technical aspects of your manuscript meet the journal's requirements, which can be a time-consuming and meticulous task.

Robots Cooking GPT

The core knowledge source for Robot Chef is the content from this blog. I carefully curated a subset of posts to form the foundation for Robot Chef's knowledge base.

When a user interacts with Robot Chef, it prompts them to select a topic of interest from a predefined list. This structured approach keeps the conversation focused while allowing users to explore areas that intrigue them. The list of topics mirrors the diverse subjects and themes covered in this blog.

Once a user selects a topic, Robot Chef can have free-flowing discussions, share relevant information, generate new ideas, or generate visualizations of the idea etc.

The key goal is to transform my static blog content into engaging, interactive conversations. Robot Chef allows readers to dig deeper into ideas I've written about by having dynamic dialogues powered by an LLM.

It is a simple experiment to fuse conversational interfaces with my own content creation. As of now I'm learning how to refine it and can see several paths forward... I'll update via this blog as I go.

I also renamed the JATS Converter GPT to "Academic Metadata Extractor"

© Adam Hyde, 2023, CC-BY-SA
Image public domain, created by MidJourney from prompts by Adam.