Google’s DeepMind creates Artificial Intelligence that can ‘Imagine’
8 months ago Global Daily Tribune . 0
Google’s DeepMind is now thinking of something big. They are now developing an AI (Artificial Intelligence) which is capable of thinking just like human beings and it can also handle the unpredictable scenarios of real life world and situations. It is also capable of seeing the consequences of their actions before they make them.
According to a report on Thursday, Google that acquired DeepMind in 2014 is developing AI (Artificial Intelligence) which can imagine like humans and see the consequences before taking any action. The report noted that its attempt to create algorithms will stimulate the distinctly human ability to construct a plan that will eventually produce software and hardware and will help in solving the complex tasks effectively and efficiently.
DeepMind was also earlier successful in AI when it developed ‘AlphaGo’ that had beaten recently the human champions at the tricky game ‘AlphaGo’. But in the case of this came, there were certain rules and predictable outcomes.
The DeepMind researcher has also quoted that “the real world does not have rules and the outcomes are also not predictable and hence it is complex. Even it is difficult for the intelligent agent to imagine these complex real life situations and to survive as it also involves a lot of costs.
A network known as the Neural network that learns to extract the information which would be helpful in future decision making can be done with the help of “imagination-augmented agents”. Here an AI agent can learn how to construct different strategies and choose the best strategy from the list of various strategies.
The US based AI (Artificial Intelligence) company and the researchers from DeepMind last month said that “there were developments that could help AI to learn about the world around it and perform the complex tasks simply, effectively and efficiently.