Pilot AISmrteasy

Mr. Know-All – 2023.7

Created: 2024-03-21

Created: 2024-03-21 14:17

I believe 2023 can be called the era of AI, more specifically, generative AI. It's not an exaggeration to say that OpenAI's Chat-GPT kicked it off.


New news keeps popping up, along with related YouTube videos.

I initially intended to organize and share them daily, but it's proven difficult, so I'm going to do it monthly instead.

I'm launching a monthly AI magazine called “Mr. Know-All”.


[Mr. Know-All Issue 1 – July 2023]

Claude 2 – a powerful new LLM has emerged. Developed by Anthropic, a company founded by individuals who were independent of OpenAI. It's said to be a great alternative to OpenAI.





This article explains how to use the embedding service in Azure to find meaningful information within unstructured documents. Since MS has invested in OpenAI, we should also keep a close eye on Azure, which is providing AI services.


LangChain & FastAPI – Private, custom ChatBot with JWT Authentication

It's in German. Refer to the GitHub repository.


What Is Chaining? | Langchain

Chaining – a core concept in LangChain. As the name suggests, it's a chain, specifically an LLM chain.

There are three types of chains – simple sequential, multiple input, and multiple output.



I need to take a look at the PDF Preview section.


Chat with your Data using LlamaIndex and OpenAI GPT-3 Collab Python Demo

LlamaIndex is used to embed personal data.

LlamaIndex is a "data framework" to help you build LLM apps. Don't just approach LlamaIndex as a simple indexing library; it's worth taking a closer look at it thoroughly at least once.


It's about 2 hours long. Below is a summary using the OpenAI API and translated with DeepL.

Dr. Shanaprakash hosts a discussion on generative AI, a major shift in platforms since the mid-90s. He will discuss how to build the next generation of Salesforce for Shopify through generative AI, Asian experiences, architecture, and how to leverage machine learning with real-time and historical data. He will also discuss data engineering, vector search embedding, prompt engineering and testing, prompt search and inference, and how to maximize performance from vectors. Furthermore, he'll discuss the importance of providing a feedback loop for AI to learn and adapt to changing language and concepts, and the need for human oversight to understand how to detect and prevent hallucinations.

Comments0