Episode notes
This one’s wild, AI can now spot cancer from a drop of blood before symptoms even show up, and it only costs $39. Also: MIT just built an AI that teaches itself better than GPT-4.1. No joke.
We’ll talk about:
- A shocking cancer-detecting AI tool being tested on real patients
- Why ChatGPT’s trash data is messing up future AI models (yep, model collapse is happening)
- Claude 4 co-writes a spicy takedown of Apple’s viral AI paper (💀)
- How to combine OpenAI + Claude + Gemini like an AI Avengers squad
- One woman’s $10K ML win in 1 week, and the trick that got her there
Keywords: Fragle, SEAL, model collapse, Claude 4, ChatGPT, Anthropic, GPT-4.1, self-improving AI, cancer detection, Claude vs Apple, AI to ...
Keywords
ChatGPTAIAnthropicGoogle Veo 3AI Agents