AlexW wrote: ↑Mon Apr 08, 2019 10:58 am
But... all you ever know is reality - how could it be any different?
All I can ever know is my experience of reality, not reality itself. This is the foundation of empiricism.
Do I know that I am hungry or am I experiencing hunger and therefore I know I am hungry.
Do I know that it's raining outside or am I experiencing rain and therefore I know it's raining.
AlexW wrote: ↑Mon Apr 08, 2019 10:58 am
Everything I say is about the map
True
The map is experience. Memory of past events.
It's raining outside. I am saying that 0.05 seconds AFTER I experienced/observed the rain.
Everything we SAY about the "now" is actually a statement about the very-recent past.
Short and long-term memory plays part here.
AlexW wrote: ↑Mon Apr 08, 2019 10:58 am
You can only say "dog" because you have learned to pattern match / extract a certain part of experience - certain shapes and colors - and label it accordingly. But reality is known all the time, no matter if pattern matching and labelling is happening or not.
Do you know what a "dog" is or do you only CLAIM to know what a dog is?
Can you really recognize "dogs" or do you only CLAIM to be able to recognize dogs?
Science can help you answer this. With a trivial experiment.
I give you a room that has 100 animals. Dogs, wolves, wild dogs, coyotes, dholes, jackals. Basically - a mixed bag of everything from the
Canis genus
If you REALLY know what a "dog" is then you should have absolutely no problem sorting the room of animals into two categories:
* Dogs
* Not dogs
This is
binary classification 101. You WILL make Type I and Type II errors. Guaranteed
The reason is simple: you are trying to map 100 animals to 2 categories (dog vs not-dog)
To do that you are necessarily discarding differences and focusing on similarities between the concepts of "animals" and "dog". It's called
Reduction.
We use
ROC CUrves to measure the precision-level of binary classifiers.
With one caveat: There are no perfect classifiers.
The implications of that is if I were to sort the room of animals and if you were to sort the room of animals we'd probably end up with different sortings.
AlexW wrote: ↑Mon Apr 08, 2019 10:58 am
True, but reality is not about concepts. Concepts inhabit their own world, they never touch reality.
"Dog" is a concept. The fact that you are making Type I and Type II errors when deciding whether something is a "dog" or "not-a-dog" is evidence to that fact.