Approximation groups

Wikipedia: "An approximation (represented by the symbol ≈) is an inexact representation of something that is still close enough to be useful."

Approximation groups are nodes, that approximate reality. They are used to cover as many real life objects as possible with as little work as possible.

High level ethical research

There are millions of products, brands and companies in the world. Current ethical research exists only on a small percentage of these. On the other hand, ethical research often covers high level topics, which deal with most of the products, brands or companies.

It is possible to take higher level ethical research topics and cover many basically unknown products, brands or companies with them.

Chocolate is a good ethical topic, since most of the world's cocoa comes from West-Africa, where child slavery exists. If there is an unknown brand of chocolate, it contains cocoa from West-Africa for over 50% chance, which is something worth mentioning! So, if we create an approximation group called "average cocoa product", anyone can add any chocolate or other cocoa product into it. If we have a cocoa product, that is known to be different from the average, it will be grouped to some other approximation group. For example "average fair trade chocolate" is another group, more detailed approximation.

This results in having untrustable information, but if it is flagged untrustable, there is not that much harm. The idea of the model is to ask for the companies to provide facts that raise the product out of the (usually) negative approximation group. "The producer has not provided information to correct this approximation."

The philosophy is that consumers buy the most ethical products, not that the consumers boycott the unethical ones. Approximation group thinking supports this.

Everything is an approximation

Philosophically speaking it is pointless to use the term "approximation group", because if we model reality, everything is an approximation. Even though companies are legally very accurate entities, we can think of them as approximations of their actions.

Limits of approximations

We can actually create an approximation group called "an average product", which would be the highest level of approximation possible. We can calculate what is an average product and how ethical it is, but that's not useful anymore. Where is the limit of useful approximation, can we say "an average chinese electronics manufacturer" and group most of the consumer electronic products in it?

Actions required

Approximation groups require that every node contains it's approximation level in it. Approximation group is not a specific node. "Average US company" is a company, even though it represents many companies.

A debate has to be held about what are the useful approximation levels.