A PORTFOLIO OF PROJECTS THAT RANGE FROM 'GENUINELY USEFUL' TO 'WHY DID I SPEND 6 MONTHS ON THIS.' MOSTLY THE LATTER. BUT I'M VERY PROUD OF ALL OF THEM IN THE WAY A PARENT IS PROUD OF A CHILD WHO TRIED THEIR BEST.
THESE FILTERS DON'T ACTUALLY WORK YET. BUT THEY LOOK NICE.
I Built a Competitor to the FDA. They Have Not Noticed.
A system optimized for saving lives rather than avoiding lawsuits. 50,000+ observational studies analyzed. Meta-analyses for 90%+ of condition-treatment pairs. Treatment rankings for 100+ conditions. The current FDA approval process takes 10+ years and costs $2.6B per drug. During that time, between 11,000 and 115,000 people die who could have been saved if the drug had been approved immediately. There are roughly 1.16 quadrillion possible drug combinations we haven't tested. At the current pace, we'll finish testing them all in approximately never. The dFDA could accelerate clinical discovery by 80X. Which, if my math is correct, means we could map the entire space of possible treatments in about 45 minutes.
416M Lives Saved. The Pentagon Has Not Returned My Calls.
Redirect 1% of global military spending to health research. Save 416 million lives. 700× more effective than current spending. I sent this proposal to everyone with the power to implement it. They have enthusiastically ignored me. When your last name is Sinn, you take validation wherever you can find it.
2M+ Visitors. Policy Unchanged. But the Charts Are Nice.
Data-driven policy analysis. Articles like 'The War on Drugs Increases Drug Deaths' and 'We Spend More on Corporate Welfare Than Social Welfare.' 2 million people read these. The policy remains exactly the same. But I made some really compelling charts and I stand by them.
Decentralized Autonomous Organization
A DAO focused on clinical research. That's a lot of technical jargon, so let me translate: imagine if Wikipedia and a clinical trial had a baby that was raised by blockchain. The result is a platform where anyone can contribute to medical research without needing permission from a committee of very tired people in lab coats.
What If We Let AI Agents Try to Fix Everything
An autonomous agent platform where AI tries to solve global problems. The agents are very enthusiastic. Their success rate is still being calculated. But they never sleep and they never complain, which is more than I can say for myself.
Another Platform. Because One Was Not Enough.
A hub for digital health intelligence. At some point I decided that having multiple platforms was a good idea. The jury is still out on this decision.
Democracy But With More Math
Aggregated Pairwise Preference Allocation. If that sounds like something from an economics paper written at 3 AM after too much coffee, you are correct. It's a mechanism for democratic resource allocation. Whether anyone will ever use it remains unclear. But the math is elegant and that has to count for something.
Instructions Nobody Asked For
Comprehensive documentation for the War on Disease initiative. I wrote a lot of words explaining how to save 416 million lives. The manual is very thorough. The implementation remains theoretical.
The Government Program Remains Classified.
The Plutonium Kidz were born of a secret government program involving the testing of plutonium exposure on human subjects. Through these experiments three normal children were thus transformed into the Plutonium Kidz! The government program, regrettably, remains classified.
50,000+ Studies Published. Chronic Illness Remains.
I built a platform to analyze what affects human health. 14 million data points. Novel causal inference methodology. The entire thing is open source. Most people do not, in fact, want to know what affects their health. They want to eat chips and watch TV. Which is fair. I would also like to eat chips and watch TV. Instead I made this.