🎙️ EP 109: GPT-5 Just Helped Prove a Quantum Theorem + DeepSeek’s Big AI Hack

AI just helped crack a math proof in quantum computing, not formatting, not grammar… real thinking. And in China, DeepSeek launched a model that slices AI costs in half, while U.S. labs are still feeding GPUs like firewood. Something big is shifting.

We’ll talk about:

• How GPT‑5 helped Scott Aaronson in a quantum complexity proof
• Why DeepSeek’s “Sparse Attention” is a direct shot at GPT’s inefficiencies
• Sora 2’s wild move — using copyrighted content until owners say no
• USC’s $1.5M ChatGPT deal and how students are actually using it
• The one takeaway that shows AI isn’t just writing... it's now reasoning

Keywords: GPT-5-Thinking, DeepSeek v3.2, Sparse Attention, QMA, OpenAI Sora, ChatGPT in schools, AI cost, academ ...