Everyday AI Podcast – An AI and ChatGPT Podcast EP 506: How Distributed Computing is Unlocking Affordable AI at Scale
14 snips
Apr 17, 2025 
Tom Curry, CEO and Co-Founder of DistributeAI, revolutionizes the world of AI through distributed computing, making advanced technology affordable for all. He discusses the challenges businesses face with rising AI demands and how to optimize unused computing resources. The conversation highlights the evolution of AI models, focusing on efficiency and adaptability. Curry also explores the promising future of edge computing and the impactful competition between open-source and proprietary models, underscoring the importance of smart scaling in AI.
AI Snips
Chapters
Transcript
Episode notes
Using Spare Compute in Business
- A medium-sized business can lend spare computing power from idle office computers at night.
- By morning, they can access powerful AI models via API from that pooled compute resource.
GPU Demand Hits Peak Limits
- Demand for GPUs skyrocketed due to AI, replacing gaming as the main driver.
- Current chip technology is reaching its limits, unable to keep pace with AI's growing compute needs.
Smaller Models Use More Tokens
- AI models are evolving with smaller size but more token processing demands.
- Chain of Thought prompting trades model size for lengthier computation to improve output quality.