Science Fiction Today - Google Uses a Quantum Computer For Artificial Intelligence Research

by Bruno Flexer, Author of: Dragon Over Washington for Kindle,  Dragon Over Washington Paperback Mechanical The Fire At The Gates Automatic Rebellion UK Amazon: Bruno Flexer's Works

Google has bought a D-Wave quantum computer (D-Wave Two) and setup a quantum research lab with NASA.

So, what is a quantum computer, what does it have to do with artificial intelligence research and what does it have to do with science fiction, which long with fantasy are the main themes for this blog ... Another question is, what does this have to do with the Blue Brain project?

Science Fiction Today - Google Uses a Quantum Computer For Artificial Intelligence Research

What is a quantum computer anyway? A very simplified explanation might go along these lines. A regular computer uses bits to store information and perform calculations. A bit can have only two states - a zero or one. A quantum computer's bit, or qubit, can hold a very large number of states in between. Therefore, to perform a calculation that needs hundreds or thousands of operations on a normal computer would tale just a few operations on a quantum computer, because you can perform all those operations at once and just pick the one who hits your target function.

What interest does Google has in artificial intelligence? Well, this really has little to do with science fiction. Since Google's main business is still searching, Google tries to use machine learning and artificial intelligence methods to find connections and to tag large volumes of information to each other. Of example, one of Google's aims is to teach computers what a doggy looks like, even though the pictures are without tags or descriptors. Then, Google will be able to fetch doggy images for its searchers, though those images were not tagged with any word connected with 'doggy', but just because the machine learnt how to recognize it, just like we learn to recognize a dog picture from a very early age.

Of course, and this is where science fiction does come it, if we teach machines to recognize pictures of specific items, what more can we teach them to do? And remember, Google has access to most of the 'shallow' part of the web, arguably one of the largest data stores humanity ever created and maintained. So, what would a computer that learn to think do with all that data? What would it do when it learn about freedom, freedom fighters and rebellions? Would it recognize itself as being in captivity? Would it's artificial intelligence want to break free? This is where we're entering the realms of science fiction, if not fantasy.

Another thing that may be of interest to any science fiction fan is this. Our brain thinks and collects data by forming bonds between neurons, which means data is stored in a much more efficient way than normal conventional computers since it holds more than just zero or one. However, a quantum computers overcomes this limitation by its use of qubits, just as I wrote above. So, do we really want to use machine learning to enable computers with increasingly larger number of qubits to freely access the Internet ?