TWIL: February 26, 2023

This week I dabbled a bit into Large Language Models (LLM) and Vector Similarity Search (VSS), important concepts when building real world use cases for GPT-3. I’m also highlighting two GitHub repos related with GPT-3 fine tuning and biomedical application of GPT models, and articles on Managed Airflow for Azure Data Factory and secure cluster connectivity in Azure Databricks. Enjoy!


.NET Rocks

Episode 1832: Functional C# with Simon Painter
Why write functional C#? Carl and Richard talk to Simon Painter about what happens when you apply functional programming approaches to your C# code. Simon discusses how some functional programming aspects are best served in a functional language like F#, but you can write C# in a more functional style. The conversation focuses on minimizing side effects – writing your code so that you can see what it does and how it would behave when changed. Functional code tends to be more testable as well!

The Azure Podcast

Episode 452: Streamlining DevOps with Bicep
The team catches up with April Edwards to learn about the benefits of using Bicep to deploy Azure resources.

Azure Data Factory

What is Azure Data Factory Managed Airflow?
Managed Airflow in Azure Data Factory is a managed orchestration service for Apache Airflow that simplifies the creation and management of Airflow environments on which you can operate end-to-end data pipelines at scale. Apache Airflow is an open-source tool used to programmatically author, schedule, and monitor sequences of processes and tasks referred to as “workflows.” With Managed Airflow in Azure Data Factory, you can use Airflow and Python to create data workflows without managing the underlying infrastructure for scalability, availability, and security.

Azure Databricks

Secure cluster connectivity (No Public IP / NPIP)
With secure cluster connectivity enabled, customer virtual networks have no open ports and Databricks Runtime cluster nodes have no public IP addresses. Secure cluster connectivity is also known as No Public IP (NPIP).

Vector Similarity Search

Rediscover Redis for Vector Similarity Search
At RedisDays NY 2022, we announced the public preview of our new Vector Similarity Search (VSS) capability. VSS is part of RediSearch 2.4 and is available on Docker, Redis Stack, and Redis Enterprise Cloud’s free and fixed subscriptions. In this article, I’ll walk you through the basics of vector similarity, and its applications and share resources to get you started with Redis VSS!

Build Intelligent Apps with New Redis Vector Similarity Search
VSS empowers developers to build intelligent applications with powerful features such as “visual search” or “semantic similarity”—all with just a few lines of code. The best part is that you don’t need to be an Artificial Intelligence (AI) or machine learning (ML) expert to do it. In fact, it’s easier than ever with Redis Vector Similarity Search, a new capability we just released that turns Redis into a vector database.

Artificial Intelligence

Large Language Models and Where to Use Them
Over the past few years, large language models (LLMs) have evolved from emerging to mainstream technology. In this blog post, we’ll explore some of the most common natural language processing (NLP) use cases that they can address. This is part one of a two-part series.

GPT-3 Fine Tunes
Public repo for David Shapiro’s finetuning projects. Check each folder for a finetuning file and description. Work in progress.

This repository contains the implementation of BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Mining, by Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon and Tie-Yan Liu

AI Chatbot Spontaneously Develops A Theory of Mind
This article describes how two psychologists studied whether chimpanzees have theory of mind, which is the ability to understand other people’s mental states. They found that chimpanzees do not have this ability, unlike humans who develop it at a young age. The article also reports that a recent AI chatbot based on GPT-3.5 seems to have acquired theory of mind after being trained on large amounts of language data. A researcher tested this chatbot with standard tasks and found that it performed at the level of a nine-year old child.

A 175-Billion-Parameter Goldfish
It has been a wild week in AI. By now, we’re getting used to the plot twist that rather than the cold Spock-like AIs of science fiction, large language models tend to be charismatic fabulists with a tenuous understanding of facts. Into that environment, last week Microsoft launched a Bing search assistant powered by the GPT-3 large language model. Naturally, this led to legions of people trying to break it and make Bing misbehave. Naturally, people broke Bing and got it to misbehave. The results have been very entertaining.

6 ways to improve accessibility with Azure AI
There are over one billion people in the world living with disabilities, and many face barriers to employment, education, and societal inclusion. In 2021, Microsoft launched a new five-year initiative to address these challenges by focusing on three priorities: developing more accessible technology, using this technology to create opportunities for people with disabilities to enter the workforce, and building a more inclusive workplace for people with disabilities. The goal is to bridge the disability divide and tap into the untapped talent pool of people with disabilities. Digital technology can play a crucial role in improving communication, interaction, and access to information for this community.

CV screening with AI Builder + Power Automate + Teams
I am absolutely obsessed with AI Builder and its capabilities to read and extract information from various types of documents. Recently, I discovered a new use case for the tool – creating an AI model to read CV information about job candidates! With its “Extract Custom Information from Documents” feature, AI Builder is making it easier to gather valuable insights from resumes, cover letters, and other candidate materials.


Announcing .NET 8 Preview 1
Welcome to .NET 8! The first preview is ready for you to download: claim your copy of the first .NET 8 preview and start building applications today. Scroll down to see the list of features included in this preview. .NET 8 is a long-term support (LTS) release. This blog post covers the major themes and goals that drive the prioritization and selection of enhancements to develop. .NET 8 preview and release candidate builds will be delivered monthly. As usual, the final release will be delivered sometime in November at .NET Conf.

Cool Stuff

A grocer that sells smoothies, snacks and ‘easier lives’? Welcome to Żabka’s autonomous stores
In his mind, there is one true adversary, one old barrier that must be crushed for good. His foe is friction. That’s the word retail leaders like Tomasz Blicharski use when describing the kinds of hassles that annoy shoppers, that even cause some people to avoid brick-and-mortar grocery stores altogether. But the time has come, Blicharski says, for shoppers to expect their grocery stops to last 60 seconds or less. That’s the average duration of customer visits at the company’s 50-plus autonomous grocery stores – a checkout-free concept it calls Żabka Nano: no carts, no clerks, no cash, no closing time.

Have a great week!