Project Polaris & Foundry Local Workshop

Microsoft recently announced a push towards greater control of the AI stack, strategizing to unite its silicon, runtimes, models, and products while reducing dependence on OpenAI.

We will explore the technologies behind it:

1. Project Polaris, Microsoft's in-house coding model (which is intended to replace GPT-4 Turbo in GitHub Copilot)
2. Foundry Local, a solution to run AI models on device.

We will discuss their technical capabilities, developer benefits, and commercial rationale.

This is a hands-on workshop. You will be running local AI models on your computer using Microsoft's Foundry Local. (Note: if you're familiar with Ollama, Foundry Local is an alternative which runs on Mac & Windows.)

Format
(15 min) Talk on Project Polaris
(15 min) Intro to Foundry Local
(60) min) Hands-on workshop on Foundry Local

Please come prepared with:

• Laptop + charger
• Internet access or an ethernet network cable
• GitHub account
• Install Visual Studio Code
• Install Foundry Local