Reportedly announced, Google DeepMind is now capable of teaching itself depending on the information it already possessed.
Founded in 2010, Google’s British Artificial Intelligence company DeepMind is now on the edge of a revolutionary invention. DeepMind is able to learn, or we can say, to teach itself based on data it already holds. This leads to a completely self-contained learning.
According to their DNC’s -Differentiable Neural Computer- anwers, it can answer questions about the relationships in a family. The researchers from DeepMind, Alexander Graves and Greg Wayne explained the whole idea as,
We wanted to test DNCs on problems that involved constructing data structures and using those data structures to answer questions. In a family tree, we showed that it could answer questions that require complex deductions. For example, even though we only described parent, child, and sibling relationships to the network, we could ask it questions like “Who is Freya’s maternal great uncle?”
We also found it possible to analyse how DNCs used their memories by visualising which locations in memory were being read by the controller to produce what answers. Conventional neural networks in our comparisons either could not store the information, or they could not learn to reason in a way that would generalise to new examples.
Let’s watch the video, here:
The DeepMind computer’s AI is smart enough to analyze the given data in order to create its own connections based on what’s stored in memory.
What makes DeepMind’s DNC system different is that they are using their own memory for keeping and processing information similar to human brain.
This could be a serious game-changer in the technological progress for robots which might turn into ordinary human-intelligences.
It is an outstanding step taken on behalf of artificial intelligence.