Pioneers of AI Andrew Ng on winning the AI race, with DJ Patil
35 snips
Nov 26, 2025 
In this insightful discussion, Andrew Ng, a pioneering AI educator and co-founder of Coursera and DeepLearning.AI, joins DJ Patil, former U.S. Chief Data Scientist, to explore the future of AI. They delve into the concept of 'agentic' workflows and how they differ from traditional prompting methods. The duo stresses the importance of coding skills alongside AI and discusses the challenges the U.S. faces in maintaining its competitive edge in the global AI landscape. They also examine the role of semiconductors and the need for trust in AI's social acceptance.
AI Snips
Chapters
Transcript
Episode notes
How 'Agentic' Became Popular
- Andrew campaigned quietly to popularize the word "agentic" to avoid binary debates.
- Marketers later plastered the term everywhere, accelerating its adoption.
Give AI The Right Context And SOPs
- When tasks need context or multimodal inputs, codify the SOP into a multi-step agentic flow.
- Focus on building the data plumbing so AI has the context a human would use.
Adopt Stepwise Agentic Workflows
- Use agentic, stepwise workflows instead of single-shot prompts for complex tasks.
- Break tasks into outline, research, draft, and critique steps to improve accuracy and utility.