
Sign up to save your podcasts
Or


Do large language models like ChatGPT or Deep Seek actually understand the world, or are they just really good at faking it?
In this episode, we dig into the debate: are LLMs building genuine “world models” or just recycling patterns from massive datasets? We’ll explore real-world AI fails, like the skin lesion detector that linked rulers to cancer, and what that means for AI in diagnosing conditions like vitiligo.
Tune in to get the real story behind AI’s so-called “understanding.”
Suggested reading:
By Yan ValleDo large language models like ChatGPT or Deep Seek actually understand the world, or are they just really good at faking it?
In this episode, we dig into the debate: are LLMs building genuine “world models” or just recycling patterns from massive datasets? We’ll explore real-world AI fails, like the skin lesion detector that linked rulers to cancer, and what that means for AI in diagnosing conditions like vitiligo.
Tune in to get the real story behind AI’s so-called “understanding.”
Suggested reading: