OpenAI-Specific Code Nixed from Semantic Kernel AI SDK -- Visual Studio Magazine

OpenAI-Specific Code Nixed from Semantic Kernel AI SDK

By David Ramel
10/13/2023

Microsoft shipped the first beta of its Semantic Kernel SDK for AI development, in the process making it service-agnostic by removing OpenAI-specific code.

Semantic Kernel acts like an AI orchestration layer in Microsoft' stack of AI models and Copilot AI assistants, providing interaction services for those constructs to work with underlying foundation models and AI infrastructure.

Officially the company's "What is Semantic Kernel?" documentation says: "Semantic Kernel is an open-source SDK that lets you easily combine AI services like OpenAI, Azure OpenAI, and Hugging Face with conventional programming languages like C# and Python. By doing so, you can create AI apps that combine the best of both worlds."

Notice the emphasis on OpenAI, the Microsoft partner that created ChatGPT and the tech powering Microsoft's AI-powered Copilot assistants being infused throughout its products and services via AI plugins.

**[Click on image for larger view.]** Semantic Kernel in the Microsoft Ecosystem *(source: Microsoft).*

While that partnership -- based on a Microsoft investment north of $10 billion -- gave Redmond a lead in the enterprise AI races, the company sought to lessen OpenAI tie-ins with the new v1.0.0 Beta1 for the .NET Semantic Kernel SDK published this week.

The beta introduces several breaking changes necessary to get the SDK into shape so the team can lock the API down for v1.0.0, one of which relates to OpenAI-specific code.

"We are changing Semantic Kernel core to be AI service agnostic," Microsoft said. "A few of the Semantic Kernel abstractions were very OpenAI specific. We have fixed this to make the Semantic Kernel AI provider agnostic."

Microsoft's Oct. 10 announcement lists some 24 breaking changes in total, requiring developers to update a large number of NuGet packages presented in a list that, for example, reflects the dev team's switch to use the term "Plugins" to replace the term "Skills" in its library names. So the list of breaking changes shows the previous name:

Microsoft.SemanticKernel.Skills.Core

the new name: Microsoft.SemanticKernel.Plugins.Core

and any notes: "Rename from Skills to Plugins."

**[Click on image for larger view.]** Semantic Kernel Breaking Name Changes *(source: Microsoft).*

"We are moving away from using the term 'Skills' and standardizing on the term 'Plugins' instead," Microsoft explained. "To avoid confusion in the future we have renamed packages and classes that used the term 'Skill' to now use 'Plugin.' Functions and properties that use 'Skill' in their names have been marked as obsolete and new equivalents provided. To learn more about our reasoning, check out the Skills to Plugins blog post."

The post also provides a list of things for developers to do in order to update their code, such as renaming functions and performing various merges and refactorings.

Further describing what's new in the first beta, Microsoft said: "We have consolidated our planner implementations into a single package and introduced more code sharing. This will make it easier for developer to create their own custom planner implementations."

The company said the purpose of the post was to outline all the changes and how to resolve them when devs uptake the initial release candidate.

About the Author

David Ramel is an editor and writer at Converge 360.

Printable Format

comments powered by Disqus

Featured

Hands On: New VS Code Insiders Build Creates Web Page from Image in Seconds

New Vision support with GitHub Copilot in the latest Visual Studio Code Insiders build takes a user-supplied mockup image and creates a web page from it in seconds, handling all the HTML and CSS.
Naive Bayes Regression Using C#

Dr. James McCaffrey from Microsoft Research presents a complete end-to-end demonstration of the naive Bayes regression technique, where the goal is to predict a single numeric value. Compared to other machine learning regression techniques, naive Bayes regression is usually less accurate, but is simple, easy to implement and customize, works on both large and small datasets, is highly interpretable, and doesn't require tuning any hyperparameters.
VS Code Copilot Previews New GPT-4o AI Code Completion Model

The 4o upgrade includes additional training on more than 275,000 high-quality public repositories in over 30 popular programming languages, said Microsoft-owned GitHub, which created the original "AI pair programmer" years ago.
Microsoft's Rust Embrace Continues with Azure SDK Beta

"Rust's strong type system and ownership model help prevent common programming errors such as null pointer dereferencing and buffer overflows, leading to more secure and stable code."
Xcode IDE from Microsoft Archrival Apple Gets Copilot AI

Just after expanding the reach of its Copilot AI coding assistant to the open-source Eclipse IDE, Microsoft showcased how it's going even further, providing details about a preview version for the Xcode IDE from archrival Apple.