

162 - AI SERIES: 10 - Understanding the Nuances of Advanced Language Models with Richard Marmorstein
May 1, 2025
Richard Marmorstein, a software developer at Hume AI, shares his AI expertise and personal projects, including crafting a custom storybook for his son. He demonstrates Hume AI's advanced text-to-speech technology and discusses its business applications. The conversation dives into the distinction between general language models and reasoning models like those from OpenAI, exploring how AI enhances coding tasks and facilitates creative content repurposing. Richard's insights on using AI tools provide a fresh perspective on leveraging technology in daily life.
AI Snips
Chapters
Transcript
Episode notes
AI-Generated Kid's Storybook
- Richard created a custom AI-generated storybook for his son to encourage soccer participation.
- He used Google AI Studio's Gemini experimental model to produce the images and story quickly.
Expressive TTS with Language Understanding
- Hume AI's text-to-speech model is expressive and customizes voice emotions deeply.
- It uses a language model to understand text meaning, not just phonetic conversion.
Use Cases for Expressive AI Voices
- Use expressive TTS for creative content like ads, podcasts, or animations.
- For businesses, combine conversational AI with expressive voices for customer interactions.