everyone’s a live player until they think too much or too little about it. orchestrate to be orchestrated. phd: functional/regulatory genomics. complex systems
this is the exact issue. only a founder will take risks like this and absorb the pain it comes with
they know that even they blow up, they've started from nothing once and they can do it again
this is why companies die when the founder leaves
i have woken unexpectedly in the middle of deep sleep and caught myself in surprising states: with thoughts that would possibly never cross my mind when i am awake
The causal network behind a phenotype — starting with disrupted 3D genome organization or altered interactions among DNA, RNA, and proteins, depending on where the mutation first acts — will differ between a perturbation that removes a gene, RNA, or protein entirely and a naturally occurring mutation that alters it.
Any structural biologist will see this as obvious.
Since models can’t navigate through good heuristic reasoning in biology, researchers at Anthropic have baked something called preregistration into Claude models: they want to preregister everything before testing a hypothesis. That way, they have a defined boundary and a null model to work with. But the null model’s definition, and the reasoning behind it, are almost always fragile.