Generative Now | AI Builders on Creating the Future Logan Kilpatrick: Building Google Gemini
11 snips
Dec 5, 2024 
Logan Kilpatrick, Senior Product Manager at Google AI Studio, discusses his journey from OpenAI to leading the Google Gemini API project. He highlights how the innovative atmosphere at Google feels akin to a startup. Logan emphasizes the importance of functionality over technical complexity in AI applications and the opportunities arising from declining model costs. He also explores the significance of specialized AI talent and the implications of Artificial General Intelligence, while considering how AI tools like Gemini can enhance user experience and business efficiency.
AI Snips
Chapters
Transcript
Episode notes
Google and OpenAI Startup Experience
- Logan Kilpatrick joined OpenAI when it was a small startup of 200 people.
- It felt like a true startup, unlike his current experience at Google, which also feels like a startup despite its large scale.
Google's AI Structure
- Google DeepMind builds the generative AI models that power Google's internal features and developer APIs.
- Google Cloud provides access to these models for both enterprise customers (Vertex AI) and individual developers (AI Studio).
Building Differentiated AI Apps
- Focus on building applications with long-term differentiation, not just wrappers around existing models.
- Model capabilities will continue to improve, so applications need to offer unique value beyond simply accessing AI models.