Frequently Asked Questions
========

If you have any questions you fail to find here, feel free to contact us at contact@monok.com

**Are these articles plagiarism?**

Both the AI generated articles, as well as the human edited version pass multiple plagiarism checks before being delivered to you 
and are no different than someone summarizing a news story with multiple sources. 
This falls under something called "observational journalism", more commonly knowns as rewriting.
Our articles systematically score better in plagiarism checks compared to articles by Yahoo, The Daily Mail and other publishers that make use of AP and Reuters articles and only make slight changes to the article. This has been shown in our own benchmark tests using Grammarly and other third-party plagiarism checkers.

**Will Google punish my website if I use these articles?**

The human written articles are no different than any other human written article found online. 
Our AI articles are indistinguishable from human written articles to a search engine. They can pass machine learning models trained to detect AI written texts, and can occasionally score better than actual human written articles as shown by `this paper`_ from The University of Brighton.

**Do you sell the same articles twice to two different users?**

No, we provide each client with their own unique content.

**How new/fresh can a story be, for the AI to write about it?**

The system demands at least 2 sources for a story, in order for the AI to summarize it. If you also want a human edit, there will be a delay that varies based on when the edit was ordered. A news story usually spans three days with various sources publishing it throughout that time, as has been measured by our research. Our system ensures the article is generated within the first day, in general the time-lag is between 30 minutes to 8 hours.

**Do you use GPT-3/OpenAI/3rd-party API’s?**

No, our company uses a proprietary method based on years of research. This method is a de-facto state-of-the-art news summarizer.

**How come you don’t use GPT-3?**

Because current machine learning models for text generation work like auto-complete, and systematically introduce random, irrelevant facts, not included in the actual story. They may sound great to read, but they are fake news by default.



.. _`this paper`: https://truthandtrustonline.com/wp-content/uploads/2019/11/paper_29_corrected.pdf