Service Design Show · July 30 · 1 hr 13 min

Why Mindless AI Is the Real Danger / Ovetta Sampson / Episode #259

Are you tired of the relentless AI marketing hype but know you can't afford to ignore it? Discover how to stop blindly applying AI and start mastering it as a malleable "design material" to build impactful, user-centered services . It’s time to stop buying into the stories big tech companies want us to believe about what Artificial Intelligence can and can’t do. Slapping AI mindlessly onto every project doesn't just create bad user experiences. It can do real harm. To set a better example for the service design industry, host Marc sits down with Ovetta Sampson. With a resume that includes Director of UX for Machine Learning at Google and AI design leadership roles at Microsoft and Capital One, Ovetta doesn't just talk theory. She runs Right AI to help teams adopt AI safely, responsibly, and effectively. In this conversation, she breaks down exactly how to understand AI's unique properties, navigate its limitations, and mold it into exactly what you need. In This Episode: The Design Material Mindset: How to stop fearing AI and start shaping it like clay to get the outcomes you want. Steering the Machine: Practical techniques for guiding generic LLMs to reliably produce high-quality, specific results. The Context Crisis: Why understanding your users' real-world context is about to become more important than ever before. When "Wrong" Becomes Dangerous: Why AI will always make mistakes, and how to identify the exact point where those errors become harmful. Breaking the Model: Why you need to start "messing around" and intentionally trying to break AI tools to build your own grounded mastery. If you found value in this episode, please hit the Subscribe button on YouTube, Spotify, or Apple Podcasts so you never miss a conversation. Share this episode with a fellow service designer who needs to hear this! As always, enjoy, and keep making a positive impact! ~ Marc [ 1. GUIDE ] 00:00 Intro and AI Sycophancy 03:15 Mindless vs. Mindful AI 06:45 Dangers of Automation Bias 10:00 The Humanizing Trap 17:45 Design as a Verb 24:30 Emergent AI Behavior 27:30 AI Confidence Levels 36:15 The Illusion of AI Memory 39:00 Principle 1 of Mindful AI 43:30 Library vs. Netflix Analogy 49:45 Designing for Failure 55:15 The Socio-Technical Challenge 01:01:45 Advice for Designers 01:06:00 Learning Resources 01:09:00 Redefining the Designer 01:11:15 Outro [ 2. LINKS ] LinkedIn - https://www.linkedin.com/in/ovettasampson/ MindfulAI Masterclass: https://www.ovetta-sampson.com/mindfulai-masterclass [ 3. CIRCLE ] Join our private community for in-house service design professionals. ⁠https://servicedesignshow.com/circle [4. FIND THE SHOW ON] Youtube ~ https://go.servicedesignshow.com/259-youtube Spotify ~ https://go.servicedesignshow.com/259-spotify Apple ~ https://go.servicedesignshow.com/259-apple Snipd ~ https://go.servicedesignshow.com/259-snipd

0:00-1:13:08

Play in podnod

transcript

No transcript — this publisher did not publish one.

show notes

It’s time to stop buying into the stories big tech companies want us to believe about what Artificial Intelligence can and can’t do. Slapping AI mindlessly onto every project doesn't just create bad user experiences. It can do real harm.

To set a better example for the service design industry, host Marc sits down with Ovetta Sampson. With a resume that includes Director of UX for Machine Learning at Google and AI design leadership roles at Microsoft and Capital One, Ovetta doesn't just talk theory. She runs Right AI to help teams adopt AI safely, responsibly, and effectively. In this conversation, she breaks down exactly how to understand AI's unique properties, navigate its limitations, and mold it into exactly what you need.

In This Episode:

The Design Material Mindset: How to stop fearing AI and start shaping it like clay to get the outcomes you want.
Steering the Machine: Practical techniques for guiding generic LLMs to reliably produce high-quality, specific results.
The Context Crisis: Why understanding your users' real-world context is about to become more important than ever before.
When "Wrong" Becomes Dangerous: Why AI will always make mistakes, and how to identify the exact point where those errors become harmful.
Breaking the Model: Why you need to start "messing around" and intentionally trying to break AI tools to build your own grounded mastery.

If you found value in this episode, please hit the Subscribe button on YouTube, Spotify, or Apple Podcasts so you never miss a conversation. Share this episode with a fellow service designer who needs to hear this!

As always, enjoy, and keep making a positive impact!

~ Marc

[ 1. GUIDE ]

00:00 Intro and AI Sycophancy

03:15 Mindless vs. Mindful AI

06:45 Dangers of Automation Bias

10:00 The Humanizing Trap

17:45 Design as a Verb

24:30 Emergent AI Behavior

27:30 AI Confidence Levels

36:15 The Illusion of AI Memory

39:00 Principle 1 of Mindful AI