Sorry for not publishing anything, but I have been sick for almost two weeks. I hope next post will be ready next week.
-TT
The World is on the brink of changes. Read this blog to understand the most important laws of nature that remain the same. Learn the wisdom of the greatest minds in history like Charles Darwin and Isaac Newton and apply them to today.
Tuesday, February 25, 2020
Tuesday, February 11, 2020
Three levels of causation
What makes you smarter than other animals? You can understand that the stimuli you receive do not represent only facts or data. The latter is the modern word for the former. You can also understand that the stimuli you receive are connected by cause-effect relationships. Your understanding of data consists of these relationships and you can act based on them. You can also imagine new events by using these relationships. No other animal can do it. The ladder of causation consists of three levels of cognitive ability.
The first level, seeing or observing
Your brain is a great pattern detection machine. It can observe and receive stimuli around it. It does that with greater efficiency than you can think. Your brain makes all kinds of associations based on the stimuli you have received in different situations in your life. It can also make good predictions based on those associations which are based on your experiences without specific reasons. The problem is that data is mostly stupid. It does not tell about cause-effect relationships. It does not tell which is the cause and which is the effect. You have to interpret and understand the data. Your brain can make good predictions to questions like ”What if I see x doing y” or ”How are stimuli related to each other?” If artificial intelligence is at this first level, it cannot function in new situations. Every new situation has to program to it by a human being. This level is all about the observed world.
The second level, doing or intervening
When in the first level you can observe things that have already happened, in the second level you can change them on purpose. You cannot understand cause-effect relationships just by observing them without interventions or smart experiments or copy somebody else´s actions. You cannot answer the question: ”What happens to the sales of iPhones if you drop the price by 40%?” if you have no observations beforehand.
Scientific experiments made in controlled conditions are second-level tools. For example, an online retailer can direct different customers to slightly different sales pages that sell the same products. Then, it can see the data about the conversion rate of both of them. Good questions at this level: ”What if you change red to blue color?” or ”What if you ban a person from doing something?” This second level of causation also makes it possible for you to create great causal models based on your observational data. This can be done even without experimentation if the cause-effect relationships are reliable enough. This level is all about an observable new world.
The third level, imagining or retrospection
Imagination can create answers to questions without data at all. For example, you can think about what had happened if you had not done anything. You can more easily understand the reasons behind certain outcomes. For example, you can think about what could have happened if you were unlucky and separate luck from skill. You can compare your observed data to an imaginary world or an imaginary outcome. You can also invent something that is currently not from this world without making any experiments. This level is all about the world that does not exist, yet. Technological developments do not happen without this level. Reaching out beyond the existing reality is not possible.
A simple example of all levels
A simple experiment is to lower the price of something for 50 percent. The first level of causation means that you cannot know what happens unless you have done it before. No statistical methods can be used to discover what happens unless you have experienced the same price reduction before. It is not the same thing to lower the price from 3$ to 1.50$ than from the previous experience: 2$ to 1$. The second level of causation demands experimentation of letting some customers have the price reduction and not giving the same price reduction to others. The second level of causation does not answer the question: What if we had reduced the price to 2$ from 3$? It is the third level question. This requires imagination.
This is all for this time,
-TT
Subscribe to:
Posts (Atom)