- The Wolf Report
- Posts
- Thoughts on knowledge creation
Thoughts on knowledge creation
That’s true for me, and important.
Now I’m determined to adjust my life activities to improve the way that I gain knowledge.
There are several ways to gain knowledge:
Collection
Copying
Conjecture
Error correction
Knowledge collection
We collect knowledge by collecting the information that the environment has preserved.
It’s as simple as that.
We collect information all the time. Right now, information is streaming through every sense. Same for you.
We don’t retain most of that information. The small amount that we do preserve is, by definition, knowledge.
Knowledge copying
Some of the information that surrounds us has already been preserved and encoded as knowledge. We live in an ocean of knowledge. The computer I’m writing on contains a vast amount of knowledge. Behind it is the internet with even more.
Most of the knowledge is not readily usable. It might be encoded and requires decoding before we can use it.
So I access it, decode it, and make my own copy.
Decoding knowledge
Knowledge is always encoded.
An array of bits is useless without a way to decode the bits into something meaningful. The bits might be decoded into letters, the letters into sentences, and the sentences might be decoded into a recipe for making coffee.
Without the decoding mechanisms, the knowledge is unavailable.
Encoding into brain and body
We humans make knowledge most useful by decoding it and then encoding it in a system that lets us make use of it. We can encode knowledge in the brain and body. We can encode knowledge in other physical systems and then encode instructions for using the system in the brain and body.
An initial brain-and-body encoding is likely to be crude and approximate. With practice and attention, people learn to make the encoding more efficient.
A machine that encodes knowledge is likely to be crude and approximate at first. With observation and iteration, the encoding mechanism can be refined and improved.
Distant knowledge
Knowledge is embedded in a substrate.
Substrates have physical locations.
If the nearest substrate for a bit of knowledge is far away, it’s not readily usable. Knowledge needs to be nearby to be useful. Knowledge can be copied to a moveable substrate, and the substrate moved near. The knowledge can be converted to information, and the information transferred then encoded locally. One can travel to knowledge.
Once the knowledge is nearby, it can be translated to make it useful.
Conjecture
David Deutsch argues that new explanatory knowledge is created by conjecture. A conjecture is a high-quality guess.
If the guess results in something useful, then the environment will preserve the guess. Knowledge has been created.
If a guess is not useful, it may be discarded. Or it may be the basis for another guess and another.
The first guess may result in something useful but crude and unrefined.
We can then guess at ways to make the initial guess more accurate, efficient, or otherwise more useful.
Error correction
Knowledge is information that the environment tends to preserve.
It doesn’t mean that the knowledge is correct. Knowledge may be erroneous.The environment might still preserve knowledge even if it’s partly or substantially wrong.
One reason is: the environment is itself built of knowledge and some may be erroneous. An erroneous environment may preserve erroneous knowledge and dismiss things that are more nearly correct.
Error correction is undervalued.
I’ll write another post on that subject.