Cremieux went from arguing against the death penalty as a teenager to becoming one of the most influential voices dissecting science online. Now with 250k+ followers on X, he’s known for exposing p-hacking, outcome switching, and selection effects that skew research.
In this conversation with Niklas, here’s what they actually talk about: why effect sizes in journals rarely match FDA data, how gene therapy’s real bottleneck is delivery not targets, and why siRNA is the most underrated modality in biotech right now.
What you’ll hear in this episode:
The heuristics he uses to spot bad research at a glance
Why published effect sizes are often 2–3x inflated compared to FDA data
Selection effects that quietly shape everything from education outcomes to clinical trials
The real bottleneck in gene therapy (delivery, not targets) and why germline bans hold back obvious wins
How pragmatic trials and IRB reform could finally fix a broken system
Why measles might come back and the case for vaccine mandates for kids
The truth about longevity research: most “breakthroughs” just stop early deaths, not extend the right tail
China’s clinical trial engine vs the slow US system
The modality he thinks is most underpriced today: siRNA
For builders in science who care more about getting it right than playing along.
More about GUEST’S work:
Explore Infinita City:
