A Role for Private AI?

As artificial intelligence (AI) platforms such as ChatGPT, Claude, Copilot, and Gemini have become accessible to more users – from government agencies to companies to individual users – all those users have become familiar with the power of AI.

These AI platforms are all based on large language models (LLMs) that ingest vast amounts of data in order to create the associations needed to answer user queries. But in part because of the amount of data involved, in part because the associations the AI platforms make are invisible to humans, the platforms are considered "black boxes," with their logic impenetrable to users. And it creates a belief that only LLMs can provide the depth of information needed for AI platforms to work.

Learning Objectives:

• Identify the challenges in building smaller-scale custom use cases, including tactical edge AI
• Evaluate the high-level performance, visibility and security considerations involved in private AI infrastructures
• Outline the factors to be considered by agencies looking to deploy their ow on-prem and edge AI infrastructure rather than using public cloud providers
• Delineate requirements for hosting private, agentic-ready AI platforms