How I use LLMs

This podcast is a monologue by Andrej Karpathy providing a comprehensive overview of Large Language Models (LLMs) and their applications in 2025. He begins by discussing the evolution of LLMs from ChatGPT's initial release to the current diverse ecosystem of providers and models, emphasizing the importance of understanding the model being used (e.g., GPT-4.0 vs. smaller versions). He then explores various LLM capabilities, including knowledge-based queries, "thinking models" for complex problem-solving, tool use (internet search, Python interpreter), and multi-modality (audio, image, video interaction). Specific examples are given throughout, such as using LLMs for travel advice, debugging code, and generating custom research reports. The podcast concludes by highlighting quality-of-life features like ChatGPT's memory function and custom GPTs for personalized tasks.