Mayo Clinic researchers just showed AI detecting pancreatic cancer on routine CT scans up to three years before clinical diagnosis — with 73% sensitivity versus 39% for human radiologists on the same scans.

In this clip from ThursdAI, the panel breaks down why this matters, why early detection is such a big deal for pancreatic cancer specifically (which has one of the lowest survival rates of any major cancer), and how AI could change healthcare in ways that go far beyond chatbots and coding assistants.

We talk about:
What the Mayo Clinic / RedMOD study actually found
Why CT scans people are already getting could become a mass screening tool
How this compares to recent University of Geneva work on colon cancer
Why a 25–30% error rate still saves lives when you're prioritizing screening
What this means for the future of AI-driven preventative medicine

This is the use of AI we've been waiting for.

⏱ Chapters
00:00 The Mayo Clinic announcement
00:35 73% vs 39% — AI vs human radiologists
01:00 Why this is "literally solving cancer"
01:40 What RedMOD actually analyzes
02:15 Steve Jobs and the case for early detection
03:00 Nisten's take from talking to doctors
04:40 Why CT scans you already get become the screening tool
05:20 The 15% survival rate problem with pancreatic cancer
06:30 Why 90% of care happens outside the hospital
07:30 How long-context LLMs help with CT data
08:30 Even with 25% errors, you still save lives
09:10 "F*** cancer" — final thoughts

🎙️ ThursdAI is your weekly live AI news show — covering breaking model releases, research, and product news every Thursday.

🔗 Subscribe to our show on Spotify: thursdai.news/spotify

🔗 Apple: thursdai.news/apple

🔗 Youtube: thursdai.news/yt

And for the full show notes and links visit

👉 sub.thursdai.news 👈