Mathematicians can use Botlzmann's formulas to create a model of how to predict human behavior.
The reason why we cannot predict human behavior completely is that we don't have enough data from that person's life. The idea is that similar people with similar backgrounds behave in similar ways. To make the needed data matrix the researchers need complete, and confirmed data about the person's genotypes and social background.
When we want to predict human behavior, we must determine what behavior we want to predict. Do we want to predict physical things, like where the person moves their hand? Or do we want to predict things, like how a person behaves in the voting situation? So, do we want to predict mental or physical things?
When the AI wants to predict things like where the boxer punches next time, the AI must know data, if a boxer is left or right-handed. Then the system can use statistics to predict, which side the punches come from. And the AI can see things like muscle tension to see which hand will rise next.
When the AI wants to predict human mental behavior, it can create a psychological profile. People normally vote for people, who have similar values to them. So, the system must collect data about a person's writings and other lifestyle. Things like outfits and other things tell about the values that person has, and then the AI can search for who has similar arguments in politics.
And of course, things like addictions like the price of cigarettes are important for people, who smoke much. Those kinds of things also drive voting behavior.
Isaac Asimov's psychohistory is based on the idea, that similar people behave in similar ways in similar situations.
The things that determine our nature are our genetic background and our experience background. If that kind of database is created for every single individual on earth, that makes it possible to create models, of how certain people behave.
The AI needs information for those profiles. The thing is that the AI can handle a very large data mass. When we think about Isaac Asimov's psychohistory, which predicts how a large human group behaves, the system uses Boltzmann's formulas to predict the gas movements. In Asimov's model, the atoms are humans. And the winds are the political movements.
But when Isaac Asimov wrote his legendary SciFi novel-series "Foundation" there were no quantum computers and AI. And the DNA was unknown. The DNA is the human genetic code, that tells who we are. When we think about the ability to predict how people behave, we can make a matrix, of how similar people acted in history. Similar people who are in similar positions with similar backgrounds act in similar ways.
Our genetic background along with our experiences determines our behavior. And in those models, the AI uses data about how similar people behave in history, and then those systems search for similar people. So, if the AI wants to predict how a certain person behaves, the AI must find a similar person with a similar background and then see, how that person behaves.
https://bigthink.com/the-present/ai-model-decision-making/
https://en.wikipedia.org/wiki/Psychohistory
Comments
Post a Comment