Uploaded June 2025 | Updated September 2026, 2 weeks ago
Explore Udacity's Agentic AI catalog here: bit.ly/4kGtWhX
In this live session, Dr. Richard Socher, Co-founder and CEO of you.com (an enterprise AI productivity platform redefining how knowledge work gets done) breaks down how we benchmark agentic AI systems, prevent error cascades in multi-agent architectures, and navigate the line between knowledge agents and action agents.
What You'll Learn:
- What makes Agentic AI accurate—and where it fails
- How to benchmark and validate multi-agent systems
- Real-world challenges of trust, autonomy, and margin of error
- The difference between knowledge agents and action agents
- How today’s top AI leaders are approaching ethical risk
- And much more…
Whether you’re building with AI or deciding when to trust it, this session will give you critical, behind-the-scenes insight from one of the field’s most respected experts. Plus, stick around for a live AMA where you can ask your biggest questions about accuracy and trust in AI systems.
About The Speaker:
Dr. Richard Socher is the co-founder and CEO of you.com and co-founder and managing partner of AIX Ventures. Richard previously served as the Chief Scientist and EVP at Salesforce. Before that, Richard was the CEO/CTO of AI startup MetaMind, acquired by Salesforce in 2016. Richard received his Ph.D. in computer science at Stanford. He is widely recognized as having brought neural networks into the field of natural language processing, inventing the most widely used word vectors, contextual vectors and prompt engineering. He is the 4th most-cited researcher in Natural Language Processing, with over 210,000 citations. He also served as an adjunct professor in the computer science department at Stanford.
Video Chapters:
00:00 Introduction to Agentic AI and Richard Socher
01:51 Understanding Agentic AI: Definition and Importance
04:18 Distinguishing Agentic AI from Traditional AI
06:37 Current Utilization of Agentic AI in the Wild
10:40 Real-World Applications of Agentic AI
12:55 Trust and Accuracy in Agentic AI Development
17:53 Key Performance Indicators for AI Agents
21:56 Benchmarking in AI: Establishing Standards
24:48 Multi-Agent Systems: Interactions and Complexities
30:09 The Evolving Turing Test
31:48 Governance and Trust in AI Agents
32:51 Regulating AI in Critical Applications
35:15 Determining Margins of Error in AI
36:23 The Complexity of AI Agent Decision-Making
40:18 Evaluating AI Agents: Challenges and Solutions
44:45 Navigating the AI Landscape: Skills and Opportunities
55:08 Building AI Solutions: Focus on Real Problems
Explore Udacity's Agentic AI catalog here: bit.ly/4kGtWhX
In this live session, Dr. Richard Socher, Co-founder and CEO of you.com (an enterprise AI productivity platform redefining how knowledge work gets done) breaks down how we benchmark agentic AI systems, prevent error cascades in multi-agent architectures, and navigate the line between knowledge agents and action agents.
What You'll Learn:
- What makes Agentic AI accurate—and where it fails
- How to benchmark and validate multi-agent systems
- Real-world challenges of trust, autonomy, and margin of error
- The difference between knowledge agents and action agents
- How today’s top AI leaders are approaching ethical risk
- And much more…
Whether you’re building with AI or deciding when to trust it, this session will give you critical, behind-the-scenes insight from one of the field’s most respected experts. Plus, stick around for a live AMA where you can ask your biggest questions about accuracy and trust in AI systems.
About The Speaker:
Dr. Richard Socher is the co-founder and CEO of you.com and co-founder and managing partner of AIX Ventures. Richard previously served as the Chief Scientist and EVP at Salesforce. Before that, Richard was the CEO/CTO of AI startup MetaMind, acquired by Salesforce in 2016. Richard received his Ph.D. in computer science at Stanford. He is widely recognized as having brought neural networks into the field of natural language processing, inventing the most widely used word vectors, contextual vectors and prompt engineering. He is the 4th most-cited researcher in Natural Language Processing, with over 210,000 citations. He also served as an adjunct professor in the computer science department at Stanford.
Video Chapters:
00:00 Introduction to Agentic AI and Richard Socher
01:51 Understanding Agentic AI: Definition and Importance
04:18 Distinguishing Agentic AI from Traditional AI
06:37 Current Utilization of Agentic AI in the Wild
10:40 Real-World Applications of Agentic AI
12:55 Trust and Accuracy in Agentic AI Development
17:53 Key Performance Indicators for AI Agents
21:56 Benchmarking in AI: Establishing Standards
24:48 Multi-Agent Systems: Interactions and Complexities
30:09 The Evolving Turing Test
31:48 Governance and Trust in AI Agents
32:51 Regulating AI in Critical Applications
35:15 Determining Margins of Error in AI
36:23 The Complexity of AI Agent Decision-Making
40:18 Evaluating AI Agents: Challenges and Solutions
44:45 Navigating the AI Landscape: Skills and Opportunities
55:08 Building AI Solutions: Focus on Real Problems
