4 ms·
Provide a mark of credibility bureaucrats can understand. They can't evaluate the merits of a paper on, say, dendritic computation, but they know what level of
by neural_thing 3y ago
Provide a mark of credibility bureaucrats can understand. They can't evaluate the merits of a paper on, say, dendritic computation, but they know what level of respect a Nature Neuroscience publication generally commands.
My solution is to build new centres of research without bureaucrats.
- godelski 3y ago> Provide a mark of credibility bureaucrats can understand. I think they accomplish the second half but have routinely established that they are incapable of doing the first half. 1) I don't think you can honestly validate or invalidate most research, though there is a bias towards being able to invalidate. I mean in science you tend to not prove things, but disprove things. People forget that the distinctions between things aren't "True vs False" but there's a third option "Indeterminate". Whether you want to point to Godel, Turing, or Schrödinger or whoever. This binarization actually harms science. 2) Metrics for metric's sake is highly destructive to our society and science. Goodhart's Law has clearly been taken to the extreme in our modern society. It's the exact reason everyone hates bureaucrats. Because there is a lack of nuance. The world has gotten more complex, removing nuance from our conversations is going in the opposite direction. Increased complexity can't result in increased hand waving. You just make more errors, at a rate proportional to the increasing complexity.