In Episode 13 (September 19, 2025), Jordan showed an interactive 3D model of the liver, intestines and blood vessels that he built entirely by chatting with Claude. LDL particles float around the body, and as you switch on Crestor, raise the dose and add Zetia and Repatha, they thin out. He didn't give an exact build time, only that it "didn't take me very long."
Jordan has genetically high cholesterol and takes several medications, but he had no picture of what they actually do or where in the body they work. Statins are among the most prescribed drugs in the US, so he figured plenty of people are in the same spot.
He saw two uses. A patient could come home from the doctor and build something like this to understand a new prescription. And a cardiologist could build these models in real time to show a patient how a treatment works. Sam admitted at 04:55 that he usually just trusts his doctor and hopes for the best, and liked the idea of a quick visual at the moment of prescribing.
I'm not writing any code at all. I'm not using any third party tools. I'm I'm simply chatting with Claude. I chose Claude because they have a very good web rendering engine.
— Jordan Metzner, Episode 13
He added that if he had done it with ChatGPT, he thinks it "wouldn't come out nearly as good." For more on that matchup, see ChatGPT vs Claude.
But, you know, if I look at my original prompt, it says, like, build me an animated web tool that shows how the different cholesterol medications work, turning on and off these different medications, and showing how they work.
— Jordan Metzner, Episode 13
It works as a teaching visual. Starting from no medication, the body is full of LDL dots. Turning on Crestor makes a slight change, raising the dose makes more, and adding Zetia and then Repatha leaves far fewer. You can watch the effect from the intestine all the way through. It's a model Jordan built to understand his own treatment, not a clinical simulation, and the episode didn't claim the numbers were medically precise.
Not much on the build itself; the lesson was about where else this works. Jordan pointed to anything biological, from birth to how a plant does photosynthesis, as a good fit for this kind of organic diagram. Sam wondered about visuals for other fields, like agriculture or coaching. Jordan also noted that most patients already Google or ask ChatGPT about their prescriptions; this goes one step further.
In Episode 30, Jordan went further with his own health data in the personal DNA and health insights dashboard. More examples are on AI in healthcare.
Jordan started with one prompt asking for an animated web tool that shows how different cholesterol medications work, with each drug toggled on and off. Claude produced a 2D model first; he asked it to scale up to 3D, then added autoplay demos and labels through back-and-forth chat.
It built Jordan a full 3D model of the liver, intestines and blood vessels that you can rotate and zoom, without him writing any code. He chose Claude for its web rendering and thought ChatGPT wouldn't come out nearly as good.
LDL particles floating through the liver and blood vessels. Turning on Crestor and raising the dose reduces them, and adding Zetia and Repatha reduces them further. It's an educational visual Jordan built for himself, not a clinical tool.
Jordan sees two users: patients who want to understand a new prescription after a doctor's visit, and physicians such as cardiologists who could build a model in real time to explain a treatment.
