3/ As an analogy, say you're optimizing over a preposterously high-dimensional state-space. To make the search faster, you distribute the effort. Once you find a good point, it's almost always useful to search the immediately surrounding space.
-
Show this thread
-
4/ But, should you dedicate all your resources there? Of course not. Yes, you almost certainly can improve your result by exploiting the local structure of the problem. But, if you fail to explore other, disparate regions, you'll converge on a mere local maximum.
1 reply 0 retweets 1 likeShow this thread -
5/ Science is like that. Scientists engage in a collective search for good explanations. When one scientist finds something interesting, others swarm around. That's part of how science works. It's an algorithm for allocating attention.
1 reply 0 retweets 1 likeShow this thread -
6/ But, scientists aren't machines. They are human beings whose social machinery and biography biases their expectations. So, when
@Nature and other scientists court diversity and representation, they don't do so in opposition to the search -- they do so in service to it.1 reply 1 retweet 2 likesShow this thread -
7/ Scientists from under-represented groups have different experiences than those from over-represented groups. These different experiences often lead them to see a problem differently. Or, to accumulate a different set of scientific tools that they can apply to the problem.
1 reply 1 retweet 5 likesShow this thread -
8/ THIS IS A GOOD THING! In that group are the explorers who are more likely to find new and fruitful points in the space of possibilities. Excluding them excludes useful information. It slows discovery.
1 reply 0 retweets 1 likeShow this thread -
9/ "But, a good scientist would recognize Truth™ even if the methods or citations or language was strange to them." No, that's not guaranteed! If every scientist looked at remarkable proffered findings, they *probably* would, collectively, recognize it for what it was.
1 reply 0 retweets 1 likeShow this thread -
10/ But, individually, they have limited time, methodological biases, and institutional demands. And, here's the problem: those things are conditioned by group identity. So, there is a strong bias against the inclusion of new and useful information.
1 reply 0 retweets 3 likesShow this thread -
11/ When you make a "purely objective" argument in favor of diversity and representation, you're accounting for human limitations and adjusting your loss function accordingly. That's good science, not YouTube celebrity science.
1 reply 0 retweets 3 likesShow this thread -
12/ Which brings me back to JBP and his acolytes. What are they defending if it's not science? The stereotype of scientists being people who look like them! Ironically, their rejection of diversity in the name of science is a defense of their identity at the expense of science.
1 reply 1 retweet 8 likesShow this thread
💥 (wannabe) Ƀreaker of (the Bad) Loops 💫 Retweeted
See also: @OmanReagan's long running thread on the vacuous, posturing JBP.
https://twitter.com/OmanReagan/status/958478386959081472 …
💥 (wannabe) Ƀreaker of (the Bad) Loops 💫 added,
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.