Common Questions & Answers

1. What is Cortext? Cortext is described as "The Re-Search Engine." It is a lightweight Large Language Model (LLM) project designed to analyze and process language in a unique way.

2. What is the mission of the Cortext project? The website states two primary missions:

• To preserve indigenous languages.

• To reduce the energy footprint of LLM AI by expanding Cortext to 20 other languages.

3. How does Cortext work? Instead of operating like larger, brute-force LLMs, Cortext is described as a "lightweight" model. It works by extracting "hashtags" from text and then searching for the most overlapping "Blue Words" (likely links) in Wikipedia. The site notes it is based on "Wierzbicka's Semantic Primes."

4. What specific tools or features does Cortext offer? The website lists several applications and demos:

• Mind Maps & Tag Clouds: Visual tools for representing text data.

• Congress Bill Videos: A tool to "Discover Implications of Bills in U.S. Congress."

• Risk Wiki: A wiki designed to outline the risk factors of publicly traded companies for managing exposure.

• Straight Shooter Index: A model that analyzes news (from 2015-2017) to find patterns in sentiment, fake news, and hearsay.

5. How is Cortext different from ChatGPT or Google's AI? The site positions Cortext as a more "elegant" and lightweight alternative. It contrasts its approach, which is based on "Semantic Primes," with "Word2Vec" style processing used by the "powerful LLMs" from large corporations.

6. Can I use Cortext myself? The site describes a "Serverless Architecture" that would allow anyone to use Cortext via a browser highlight, phone app, or extension to generate reports, parse events, and stash research.

7. Who is behind this project? The contact email listed is jefferson@richards.plus, which connects it to the other "Richards" websites.

8. How can I support this project? The website includes a "Donate or Sponsor" section for those interested in supporting its mission.