MENGENAI EPISOD INI
Join host Don as he sits down with AI veteran Uvika Sharma (discussed as Iva/Yuva) on The Human Code to explore the vital intersection of humanity, technology, and responsible AI. With over 23 years of experience in data and artificial intelligence, Uvika discusses why technology transformations are ultimately "people problems" rather than software failures.
In this episode, they dive deep into the necessity of change management, the dangers of blindly trusting probabilistic AI, and the critical role of governance and "human-in-the-loop" systems. Learn how teams can use Agentic AI workflows responsibly, hear real-world (and humorous) horror stories of AI interpretation mishaps, and discover what the future holds for emerging technologies in solving global issues like accessibility and poverty. Finally, Uvika shares invaluable advice for the next generation on keeping their curiosity, critical thinking, and empathy alive.
Table of Contents
Introduction to the Human Code (00:00 - 00:54)
The episode opens with thoughts on validating AI-generated output for accuracy. Host Don introduces guest Uvika Sharma to discuss the intersection of human cognition and artificial intelligence.
Uvika’s Journey into Responsible AI (00:54 - 01:40)
Uvika shares her background of over 23 years in the data and AI space, and what motivated her to focus on the humanity aspect of tech adoption.
AI Adoption as a Change Management Challenge (01:40 - 03:10)
An exploration of why technological transformations are actually "people problems" and why companies must invest in change management for AI to succeed.
Cognition, Speed, and the "Verify" Muscle (03:10 - 08:04)
A discussion on how AI impacts human cognition. Uvika and Don stress the importance of actively questioning and validating probabilistic results instead of prioritizing speed over accuracy.
Expanding Expertise Safely with T-Shaped Skills (08:04 - 10:00)
Don explains the concept of hiring "T-shaped" individuals and how AI can massively expand a person's breadth of work, provided they stay curious and spot-check outputs.
Success Patterns: AI Workflows vs. Replacing People (10:00 - 13:28)
Uvika shares insights on how successful organizations utilize AI to assist with repetitive pain points in workflows while keeping human oversight, rather than attempting full replacement.
Managing Agentic AI and Implementing Governance (13:28 - 16:24)
Don shares a case study of 5 engineers using agentic workflows to produce the output of 50. Uvika relates managing these probabilistic systems to "baby-proofing" a house for a toddler.
AI Interpretation Mishaps & Horror Stories (16:24 - 19:30)
The hosts share cautionary, real-life examples of AI taking instructions too literally, from turning off critical product notifications to an inbox-cleaning AI deleting a user's entire email archive.
What Excites Uvika About the Future of Tech (19:30 - 23:32)
A look at the positive potential of AI over the coming months, specifically its role in medical breakthroughs, helping the visually impaired with accessibility, and potentially addressing global poverty.
Life Advice for the Next Generation: Be Positive, Kind, and Human (23:32 - 26:10)
Uvika offers her parenting and life advice for navigating an AI-driven future: maintain curiosity, rely on personal critical thinking, take ownership of decisions, and lead with empathy.
Sponsored by FINdustries
Hosted by Don Finley
