Knowledge modelling, part 4

Summary
Computers typically deal with calculations, character manipulation, and logical operations, applied to numbers and characters that form data and information. Computers typically do this via serial processing, step by step.

Humans typically deal with real-world pattern recognition applied to sensory input, involving knowledge about a huge number of patterns.

These two approaches have complementary strengths and weaknesses. Computers are excellent at data and information processing, but poor at handling knowledge; humans can easily handle knowledge, but are poor at handling data and information.

One key feature of knowledge is easily overlooked. Knowledge about the world is by necessity incomplete and often imperfect. Usually, there is no perfect solution to a problem that is guaranteed to be correct. This means humans will not get everything right, because in the real world, there isn’t a way to get everything right.

Another key feature of the human brain is that it works in ways that are very different from formal logic. Its massively parallel structure results in different types of memory, which support different types of knowledge. This is why we can be skilled in something without being able to say how we perform that skill, and why we can remember something in one context that we can’t remember in a different context even if our life literally depends on it.

Explicit knowledge can be accessed relatively easily via familiar methods such as interviews, questionnaires, and focus groups. Semi-tacit and tacit knowledge is often missed by these methods and requires ones that are less widely known.

These themes recur repeatedly throughout this book. The underlying mechanisms are simple, and make sense once you know about them, but the implications are far-reaching, and often the opposite of what most people would expect.

[End of this extract from the Knowledge Modelling Handbook]