This AI Tool Creates Audio From Your Interests — Without You Prompting
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode of The Next Wave, host Nathan Lands sits down with Raiza Martin, one of the creators behind Google's NotebookLM and now co-founder of Huxe, a startup building truly personalized AI. Martin shares her journey from leading innovation at Google to building Huxe, an app that generates personalized audio content—essentially "AI radio made just for you"—based on your daily habits and interests without requiring user prompts. (00:33) The conversation explores the challenges of creating AI that proactively serves users rather than waiting for commands, the importance of trust in AI products, and what it means to design AI for real human needs rather than just technological capabilities.
- The main theme centers on the evolution from reactive AI tools to proactive personal AI that initiates interactions and seamlessly integrates into daily routines
Speakers
Raiza Martin
Raiza Martin is co-founder of Huxe and was instrumental in building and shaping Google's NotebookLM, one of the most innovative AI products of recent years. She worked at Google for several years leading product innovation before leaving to start Huxe with her former Google colleagues. Martin has extensive experience in AI product development and is passionate about creating technology that fits seamlessly into human routines rather than requiring users to adapt to new interfaces.
Nathan Lands
Nathan Lands is the host of The Next Wave podcast and founder of Lore.com. He focuses on exploring cutting-edge AI developments and interviewing founders building the next generation of AI products. Lands has experience in business development and is actively building his own AI-focused venture while documenting the evolution of artificial intelligence through his podcast and newsletter.
Key Takeaways
Embrace the Power of Proactive AI Design
Martin emphasizes a fundamental shift from reactive to proactive AI, where the technology initiates interactions rather than waiting for user prompts. (04:59) She describes this as asking "how come you always have to go first?" when interacting with AI. Traditional AI tools require users to have specific intentions and goals before engaging, but proactive AI can anticipate needs and provide value without explicit requests. This represents a paradigm shift from "pulling value out of the tool" to being "delighted" or "getting ready for the day." The key insight is that successful AI products of the future will need to understand users deeply enough to provide relevant, timely assistance without being asked.
Build Around Existing Human Habits, Not Technology
Rather than starting with impressive technology and finding uses for it, Martin advocates building around what people are already doing in their daily routines. (06:51) She discovered that Huxe users consistently mentioned using the app during specific activities like driving, brushing teeth, or morning routines. This insight revealed that successful AI products should "hook into" existing habits rather than requiring users to develop new behaviors. The lesson is to identify natural moments in people's lives where AI can add value seamlessly, rather than forcing users to adapt to new interfaces or workflows.
Focus on Trust-Building Through Simple, Reliable Tasks
Martin explains that earning user trust is crucial before AI can take on more complex, autonomous tasks. (14:14) She notes that while people discuss AI automation extensively, most users currently find value in basic functions like summarization, catching up, and explanations rather than having AI perform actions on their behalf. The strategy is to excel at these foundational tasks first, making them so reliable and useful that they become normal parts of people's routines. Only after establishing this trust can AI systems progress to more autonomous functions like responding to emails or making payments.
Prioritize Helpfulness Over Manipulation
When developing AI that makes proactive suggestions, Martin emphasizes considering whether users would have taken those actions anyway without the AI's intervention. (15:15) This approach ensures the AI is genuinely helpful rather than manipulative. The key is distinguishing between suggestions that serve the user's actual interests versus those that might extract more value from the user. This principle becomes especially important as AI systems gain more insight into personal data and behavior patterns, requiring developers to maintain ethical standards in how they influence user decisions.
Invest in Uniquely Human Skills for Future Success
When asked about preparing children for an AI-driven future, Martin advocates focusing on uniquely human experiences and skills rather than just technical capabilities. (29:42) She recommends activities like watching movies, playing sports, learning musical instruments, looking at art, and reading—essentially anything that develops emotional intelligence, cultural understanding, and human connection. Her reasoning is that these experiences create the foundation for meaningful interpersonal relationships and unique perspectives that AI cannot replicate, making individuals more valuable in a world where technical skills become increasingly automated.
Statistics & Facts
- Huxe's Discord community has grown to 6,000 active members who help provide feedback and support for the product development. (36:26) Martin describes this as having "five folks on our team and an extended crew of 6,000 more."
- Martin and her co-founders spent approximately two and a half years building NotebookLM at Google before it became widely successful and gained mainstream attention.
- When interviewing early Huxe users, Martin spoke to nine people in a row who all mentioned using the app while brushing their teeth, revealing unexpected usage patterns around daily routines. (06:51)
