News
Microsoft previews the Semantic Kernel SDK for adding AI to apps
Microsoft this week announced Semantic Kernel, a new open-source framework on GitHub in early preview designed to help developers leverage artificial intelligence (AI) and big language models in their applications.
Semantic Kernel is a lightweight software development kit (SDK) that makes it easier to use things like OpenAI’s ChatGPT natural language chat responses in apps, according to John Maeda, Microsoft’s vice president of design and artificial intelligence, in a free LinkedIn Learning introductory video . It works with traditional programming languages, with current preview support for “C# and Python,” the announcement says. Microsoft plans to add language support for “TypeScript and other languages” as well.
Developers can add new “experiences” to their apps, e.g. “Bringing together a long chat exchange, flagging an important ‘next step’ that was added to your to-do list via Microsoft Graph, or planning an entire vacation instead of just reserving a seat on a plane,” explained Maeda.
Semantic Kernel, which got its start as an “internal incubation project at Microsoft”, currently supports OpenAI’s GPT-4 model and the Azure OpenAI service. Its capabilities can be “embedded in any kind of application” and “connected to external data sources and services”. Semantic Kernel is also notable for allowing developers to use complex prompts, which can be problematic when working with AI, according to Maeda in the LinkedIn Learning video:
So imagine being able to easily create complex prompts, multi-faceted prompts, prompts that do all sorts of things, like connect to native code. i know you like that I know we like that. With Symantec Kernel, such things are free.
In this document, Microsoft described the Semantic Kernel as a “kit of parts that work together”. There is an “ASK” part that represents the user’s goal. A kernel part orchestrates the ASK, which is done via a “scheduler” part that breaks the ASK into steps for the available “resource” part. The resource part consists of the “available skills, memories and connectors”. Next, “steps” are executed and the results are sent back to the user as a “GET” via a “pipeline” part.
The Semantic Kernel SDK is currently in early preview and will likely continue to evolve, according to Microsoft’s document.
“With new breakthroughs in LLM AIs landing every day, consider this SDK a work-in-progress as the team’s best-guess practices for adding semantically rich AI to existing apps,” reads it in the document.
Microsoft’s GitHub description described Semantic Kernel as “an early alpha” release that will change rapidly along with AI developments. The GitHub page has some model code samples for developers to try. There is a “simple chat summary” feature for apps. A Book Creator code sample allows apps to create book sketches. There is also a basic connector code pattern that can be used to “authenticate and connect to an API”.
About the author
Kurt Mackie is Senior News Producer for 1105 Media’s Converge360 group.
