Welcome to our latest podcast episode on the What’s AI show! I’m your host, Louis-François Bouchard, and today we have an exciting interview with Logan Kilpatrick, an experienced developer and developer advocate at OpenAI. We delve into fascinating insights on AI, large language models, and explore Logan’s role in developer relations.
🎓 Logan’s journey in the field of AI is remarkable, with a master’s degree from Harvard and now pursuing a PhD in Applied Machine Learning. He is truly passionate about utilizing ML to tackle real-world problems and has made significant contributions to the Julia programming ecosystem.
To gain a deeper understanding of large language models (LLMs), we start the interview with a fun segment covering the technology behind models like ChatGPT. We explore the concept of Generative Pre-trained Transformers (GPT) and unravel the significance of vocabulary, tokens, prompts, and alignment within these systems.
🔧 One of Logan’s primary responsibilities at OpenAI revolves around developer relations. He works tirelessly to ensure the success of developers utilizing OpenAI’s API and ChatGPT’s plugin ecosystem. With the rapid advancements in AI, Logan emphasizes the importance of implementing state-of-the-art models and transforming them into practical, usable products.
💡 If you’re an AI enthusiast, a developer, or simply curious about machine learning, this interview is a goldmine of information. We touch upon a wide range of topics, including AI, LLMs, and OpenAI. But don’t stop here—tune in to the full interview to uncover even more valuable insights and gain inspiration for your own endeavors in the world of AI.
Listen to the whole interview on YouTube or Spotify!
FAQ
What does the interview with Logan Kilpatrick cover?
It discusses building with LLM APIs, ChatGPT, developer relations, plugins, and working with OpenAI's developer community.
What does a developer advocate do?
A developer advocate helps builders understand a platform, communicates their problems internally, and improves documentation and examples.
Why were ChatGPT plugins important to developers?
They showed how a conversational model could call external services and work with information beyond its training.
What should builders learn before using an LLM API?
Understand request structure, model limits, privacy, cost, evaluation, and how failures affect the surrounding product.
Why listen to the full podcast episode?
The longer conversation provides context on Logan's role, developer challenges, and practical lessons that a short summary cannot cover.