Great scientists and thinkers throughout the history of our humanity have managed to unravel those mysteries that seemed hidden, some like Isaac Newton or Albert Einstein coming to mind.
In fact, these great physicists and thinkers have helped us explain not only the origin and functioning of our planet, but they have also helped us accurately predict the movement of other planets, the Moon and different celestial objects.
And now a group of researchers have successfully trained a artificial intelligence so that you think like great scientists and researchers like those mentioned Albert Einstein or Isaac Newton.
This has been achieved by scientists from the Forschungszentrum Jülich who, to train the AI, used a neural network to map complex behaviors with a certain precision and all in a simpler system.
Then artificial intelligence did all this by simplifying complex interactions between various components of the system.
The team used a simplified system to create an inverse map with this already trained artificial intelligence. This system, as it went from simple to more complex components, was able to develop different theories.
What they basically wanted with this behavior, is that artificial intelligence was similar to a real physicist when it came to capturing different approaches.
To publicize this new model, they first wanted to make it clear how their artificial intelligence model differs from others that we already know on the market.
To do this, they investigated how the smallest structures of different handwriting are made up of interactions between pixels. So with the help of this AI they were able to theorize about the brightest groups of pixels that contributed to the shape of the handwritten number.
Thanks to artificial intelligence, it was possible to study a large number of interactions at the same time. So the difference between this system and others is how it is trained.
In general these models learn the theory from the data used to train them. Instead of using the general approach, they have extracted the theory learned by the computer and the language it uses to explain the interactions between the system’s components.
This can then be used to build a link between how AI works and theories that humans can understand.
