# Centum Rapid Test FY2 S3 V1

# Introduction

There is no need to wait until we finish developing technology to test whether the long text format has promise. Immediately, compile a small set of documents to simulate the product experience then either make changes or call off the entire investment before starting to develop the technology. 

# Thesis

If I manually prepare long text documents on manually selected topics, then I will be able to spend hours consuming content for fun, and so will my contacts to whom I invite to this test. 

# Process

1. Set up a new Notion workspace. 
2. Create a taxonomy of topics. Limit to 100 topics. 
3. Manually generate documents on the topics. Limit to 10K documents, but prepare at least 1K documents. Generate by finding potential links in documents, then generating documents on those linked topics. 
4. Publish Notion site. 
5. Consider using a layer on top for monitoring. 

NOTE: Using Notion will let us simulate the content, but not the possible application features. This is okay, because app features are merely auxiliary in comparison to the content which will be the critical factor. 

After the test is finished, import Taxonomy database and Documents database for record. 

# Ending

I am ending ending this test (Centum Rapid Test FY2 S3 V1) before completion in order to expedite the next iteration of this test (Centum Rapid Test FY2 S3 V2). 

Original plan was to manually write several documents (approximately 1K to 10K documents across 100 classes in the Topics Taxonomy), using methods such as 1) manually selecting topics, 2) importing information on the topics from selected sites, then 3) translating using ChatGPT. The target was to test whether “long form text can be fun”. 

However, since making the original plan, my thinking changed in the following ways: 

1. It is trivial to test whether long form text format can be fun. I have exhibited this behavior for years. Even if I were to carry out the Centum Rapid Test FY2 S3 V1 as planned, I will suffer confirmation bias. 
2. The big bet is on technology. Specifically, I need that long form text that we can generate using our AI technology will be fun for general consumers. 

We will proceed to test, instead of whether ”biased contacts react positively to manually-generated long text documents”, whether “strangers will react positively to automatically-generated long text documents”. 

# Artifacts

[Discovery Front Page for Centum Rapid Test FY2 S3 V1](Centum%20Rapid%20Test%20FY2%20S3%20V1/Discovery%20Front%20Page%20for%20Centum%20Rapid%20Test%20FY2%20S3%20%206c6db09e86d34a9ba00fb7a9f56d47c5.md)

[Focus Topic Taxonomy for Centum Rapid Test FY2 S3 V1](Centum%20Rapid%20Test%20FY2%20S3%20V1/Focus%20Topic%20Taxonomy%20for%20Centum%20Rapid%20Test%20FY2%20S3%20%201728d9b350f34daf8d845157f2e6c442.csv)