Artificial Intelligence, Alphabet's DeepMind can now adapt and teach itself things without any human intervention. It can do so with the help of the extensive amount of database at its disposal.

DeepMind uses the current data storage, combines it with its Artificial Neural Networks and the end result is the parsing of information in units it can easily understand and then learn.

Like Us on Facebook

Researchers on the DeepMind project, Alexander Graves and Greg Wayne said, "These models can learn from examples like neural networks, but they can also store complex data like computers." 

The process that takes place underneath the DeepMind AI is much like the workings of the human brain. The neural network uses a type of interconnected nodes which stimulates the specific areas which releases only give a certain type of output.

Over time, the information it collects from learning of large data dumps would enable it to find more relevant solutions and in a much smaller time span.

The DeepMind team further explained by providing an example - The system was given the basics of the London Underground public transportation system and immediately went to work finding additional routes and the complicated relationship between routes on its own.

Adapting to situations means that DeepMind can also remember past problems it has been asked to solve and also learn from its answer. So instead of finding new possible outcome, DeepMind can just derive the solution from prior experience it has had.

Earlier this year DeepMind made headlines when it defeated a human in a game of "Go", which is known to have an infinite number of moves and solutions.