2025 - 02 February
Science is a strong-link problem
A very simple idea, immediately appealing. Even if you have quibbles, like “you need some weak-link work to make sure your field isn’t built on junk”, the broad direction is surely true.
The article mentions
Here’s the crazy thing: most people treat science like it’s a weak-link problem.
I would go further to say that there’s not just a tendency, but a desire by many for science to be a weak-link problem.
When I read this part:
Peer reviewing publications and grant proposals, for example, is a massive weak-link intervention. We spend ~15,000 collective years of effort every year trying to prevent bad research from being published.
…I couldn’t help but think of my current ongoing gripe with the review system, where my “high variance” paper, which is not going to revolutionize the field, but is surely publishable, is being bounced from journal to journal by reviews fitting their own expectations onto what it should be doing and saying “why didn’t you look at this instead?”
Not a big podcast person anymore, but I enjoy Ken Liu’s books enough to make this one worth a listen.