How swarm intelligence solves routing problems in 20 seconds without training data

Описание: Fred Gertz applies ant colony algorithms to solve NP-hard optimization problems in 20 seconds without training data. While LLMs struggle with complex scheduling and routing, his swarm intelligence approach combines search and optimization in one pass, making advanced operations research accessible to manufacturers running on spreadsheets.
In this episode, Fred breaks down why Six Sigma's focus on variation reduction misses the point, how Taguchi methods map process parameters to business outcomes, and why agent coordination across 500+ enterprise agents is swarm intelligence's next frontier.
CHAPTERS
00:00 Introduction
00:14 Fred's background: From nanotechnology PhD to data science
04:47 What is swarm intelligence and how it differs from traditional AI
08:12 Solving data variety problems in manufacturing
14:53 Why Six Sigma misses the point: Taguchi vs variation reduction
20:30 Operations research: Making complex math accessible
24:14 Deploying AI solutions in low-data environments
29:14 What customers are thinking about: AI projects and transformation
32:00 Brand voice and personality evolution at Collide
35:01 Validation frameworks: IQ/OQ/PQ for AI in regulated industries
38:28 Agent coordination: The next frontier for swarm intelligence
41:24 Predictions for 2026: LLM plateau and optimization's rise