Understanding AI Risks with Peter Slattery

In this episode of the Behavioral Design Podcast, hosts Aline and Samuel are joined by Peter Slattery, behavioral scientist and lead researcher at MIT’s FutureTech lab, where he spearheads the groundbreaking AI Risk Repository project. Together, they dive into the complex and often overlooked risks of artificial intelligence—ranging from misinformation and malicious use to systemic failures and existential threats.

Peter shares the intellectual and emotional journey behind categorizing over 1,000 documented AI risks, how his team built a risk taxonomy from 17,000+ sources, and why shared understanding and behavioral science are critical for navigating the future of AI.

This one is a must-listen for anyone curious about AI safety, behavioral science, and the future of technology that’s moving faster than most of us can track.

--

LINKS:

• Peter's LinkedIn Profile
• MIT FutureTech Lab: futuretech.mit.edu
• AI Risk Repository

--

Interesting in collaborating with Nuance? If you’d like to become one of our special projects, email us at hello@nuancebehavior.com or book a call directly on our website: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠nuancebehavior.com.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Support the podcast by joining ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly Pro⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ 🚀. Members get access to extensive content databases, calls with field leaders, exclusive offers and discounts, and so much more.

Every Monday our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly newsletter⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ shares the best articles, videos, podcasts, and exclusive premium content from the world of behavioral science and business.

Get in touch via ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠podcast@habitweekly.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

The song used is ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Murgatroyd by David Pizarro⁠

Enter the AI Lab: Insights from LinkedIn Polls and AI Literature Reviews

In this episode of the Behavioral Design Podcast, hosts Samuel Salzer and Aline Holzwarth explore how AI is shaping behavioral design processes—from discovery to testing. They revisit insights from past LinkedIn polls, analyzing audience perspectives on which phases of behavioral design are best suited for AI augmentation and where human expertise remains crucial.

The discussion then shifts to AI-driven literature reviews, comparing the effectiveness of various AI tools for synthesizing research. Samuel and Aline assess the strengths and weaknesses of different platforms, diving into key performance metrics like quality, speed, and cost, and debating the risks of over-reliance on AI-generated research without human oversight.

The episode also introduces Nuance’s AI Lab, highlighting upcoming projects focused on AI-driven behavioral science innovations. The conversation concludes with a Behavioral Redesign series case study on Peloton, offering a fresh take on how AI and behavioral insights can reshape product experiences.

If you're interested in the intersection of AI, behavioral science, and research methodologies, this episode is packed with insights on where AI is excelling—and where caution is needed.

LINKS:

• Nuance AI Lab: Website

TIMESTAMPS:
00:00 Introduction and Recap of Last Year's AI Polls
06:27 AI's Strengths in Literature Review
15:12 Emerging AI Tools for Research
19:31 Evaluating AI Tools for Literature Reviews
23:57 Comparing Chinese and American AI Tools
26:01 Evaluating Literature Review Outputs
28:12 Critical Analysis and Human Oversight
35:19 The Worst Performing Model
37:21 Introducing Nuance's AI Lab
38:51 Behavioral Redesign Series: Peloton Example
45:21 Podcast Highlights and Future Guests

--

Interesting in collaborating with Nuance? If you’d like to become one of our special projects, email us at hello@nuancebehavior.com or book a call directly on our website: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠nuancebehavior.com.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Support the podcast by joining ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly Pro⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ 🚀. Members get access to extensive content databases, calls with field leaders, exclusive offers and discounts, and so much more.

Every Monday our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly newsletter⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ shares the best articles, videos, podcasts, and exclusive premium content from the world of behavioral science and business.

Get in touch via ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠podcast@habitweekly.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

The song used is ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Murgatroyd by David Pizarro⁠

When to AI, and When Not to AI with Eric Hekler

"People are different. Context matters. Things change."

In this episode of the Behavioral Design Podcast, Aline is joined by Eric Hekler, professor at UC San Diego, to explore the nuances of AI in behavioral science and health interventions. Eric’s mantra—emphasizing the importance of individual differences, context, and change—serves as a foundation for the conversation as they discuss when AI enhances behavioral interventions and when human judgment is indispensable.

The discussion explores just-in-time adaptive interventions (JITAI), the efficiency trap of AI, and the jagged frontier of AI adoption—where machine learning excels and where it falls short. Eric shares his expertise on control systems engineering, human-AI collaboration, and the real-world challenges of scaling adaptive health interventions. The episode also explores teachable moments, the importance of domain knowledge, and the need for AI to support rather than replace human decision-making.

The conversation wraps up with a quickfire round, where Eric debates AI’s role in health coaching, mental health interventions, and optimizing human routines.

LINKS:

• Eric Hekler:

TIMESTAMPS:
02:01 Introduction and Correction
05:21 The Efficiency Trap of AI
08:02 Human-AI Collaboration
11:04 Conversation with Eric Hekler
14:12 Just-in-Time Adaptive Interventions
15:19 System Identification Experiment
28:27 Control Systems vs. Machine Learning
39:44 Challenges with Classical Machine Learning
43:16 Translating Research to Real-World Applications
49:49 Community-Based Research and Context Matters
59:46 Quickfire Round: To AI or Not to AI
01:08:27 Final Thoughts on AI and Human Evolution

--

Interesting in collaborating with Nuance? If you’d like to become one of our special projects, email us at hello@nuancebehavior.com or book a call directly on our website: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠nuancebehavior.com.⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Support the podcast by joining ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly Pro⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ 🚀. Members get access to extensive content databases, calls with field leaders, exclusive offers and discounts, and so much more.

Every Monday our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Habit Weekly newsletter⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ shares the best articles, videos, podcasts, and exclusive premium content from the world of behavioral science and business.

Get in touch via ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠podcast@habitweekly.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

The song used is ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Murgatroyd by David Pizarro⁠