Home / Magazine
Magazine
Essays, analysis and field notes. Some pieces answer a question, some find the better question, and some put a disagreement on the table and leave it there.
Can a model produce a hypothesis, or only a sentence that looks like one?
Generative systems are fluent in the grammar of science. Fluency is not the same as a claim about the world, and the difference matters most exactly where it is hardest to see.
Does AI make science more democratic, or widen the gap?
The same tools that let a small lab punch above its weight also concentrate advantage where the compute is. Both stories are true; the question is which dominates.
What peer review actually catches
A working reviewer’s account of the errors that get through, the ones that do not, and why the two categories are not what most people assume.
When a result cannot be explained, is it still science?
Explanation has never been a formal requirement of a scientific claim. It has been a practical one, and machine learning is the first thing to test whether the distinction holds.
Where the productivity actually is
AI speeds up literature triage and code. It does not speed up deciding what is worth asking. Confusing the two produces the illusion of a faster lab.
“The boring part is the part that matters”
A computational biologist on tedium, dead ends, and why the interesting result is usually the one that survives your attempt to kill it.
Nothing matches that yet. Try a broader filter.