Some Thoughts on Filters, from a Frigid Office
I bring a jacket or sweater to work every day, despite it being 90* Fahrenheit outside, because the office is cold as hell.
I asked a colleague if he thought it was cold in the office, and he said No, not really. Another colleague responded, Yes, not only is it cold, but it is legitimately freezing.
Around 2:00pm every day, a small delegation from the "It's Cold" camp quietly takes their laptops downstairs to the lounge, where the climate is more bearable. During the morning and early afternoon, it's not too bad, but as the day drags on, the temperature continues to drag further down.
Strangely, the weather patterns of the office help me stay focused. Anything I need to get done, I try to get done by 2pm. After that, I focus on responding to emails, running errands, taking meetings or walks outside, or reading. But I still wonder why some of us find the office frigid while others barely notice.
My mind wanders to the experiments of Dutch wild-man Wim Hof. He once summited Mount Everest wearing shorts, which surprised even his cold-hardened sherpa. In another experiment, he was submerged neck deep in ice water for hours; Wim's body temperature remained the same, as did his heart rate. As a followup, to show that this "skill" was not specific to him, he trained several people to be able to accomplish the same feat (which they successfully did).
Well, this forces me to re-imagine my local "coldness" predicament as such:
Particular arrangements of particles -> Particles hit my skin -> My brain and body process -> I experience "Cold" (while my friend experiences "Mild", while another friend experiences "Freezing")
Through this lens, my "coldness" is less a statement of reality and moreso a statement on how I feel about it.
There is a powerful concept in statistics named Bayes Theorem. An attentive college undergrad can recite the mathematical/axiomatic definition of it, but this kind of obfuscates a profound point:
P(A|B) = P(B|A)*P(A) / P(B|A)*P(A) + P(B|~A)*P(~A)
In less technical terms: the probability of one event given the occurrence of another event depends just as much on what you don't see as it does on what you do see.
The canonical example describes the following scenario:
- 1% of women have breast cancer.
- 80% of mammograms detect breast cancer when it is there.
- 9.6% of mammograms detect breast cancer when it is NOT there.
Suppose a woman takes the mammogram and gets a positive result. What are the chances that she actually has breast cancer? 80%?
Instinctively us non-Bayesians might think "Hey, well she has breast cancer, and she took a mammogram and tested positive, so it's gotta be an 80% chance of being correct."
What we fail to take into account is that false positives and false negatives skew our results more than we think.
Our Perceptions Are Shitty Filters
More generally, our perceptions are "mammograms" -- they are tests on reality. We see things that aren't really there and we miss things that are there, which is how our perceptions deceive us on whether there really is a there there.
A modern American primate will feel cold and adjust the thermostat (and therefore the environment) to suit his personal Bayesian filter for what Cold is.
A hardier German primate will feel Cold and put on a warm jacket.
And a highly adaptive Himalayan primate will blink at you in confusion: "Cold?"
When identifying, considering, and solving problems, some require global environmental change (thermostat), some require local environmental change (jacket), and some require internal change (adaptation). In each of these cases, we are forced to deal with our filters, determine how accurately they represent reality, and then figure out what to do from there.
So I begrudgingly put on a warmer jacket.