LLMs can get brain rot too
This episode explores the impact of low-quality pre-training data on large language model capabilities. It also covers the percentage of LLM-generated content online, the cost of embedding models, and new announcements from OpenAI and Amazon related to AI.