Google Files Patents for Artificial Intelligence

/ 1 year ago

artificial intelligence

Google has filed six patents for artificial intelligence and neural networks, it has been revealed. It’s the first time that the tech giant has attempted to protect its AI research, some of which could be spurious, while as a whole could be seriously detrimental to any future AI research and development by smaller companies.

The first patent is for what is known as dropout, a method for neural network learning, invented by Alexander Krizhevsky, Ilya Sutskever, and Nitish Srivastva of Toronto University, but used as a standard technique by most AI researchers. According to the patent, dropout is:

“A system for training a neural network. A switch is linked to feature detectors in at least some of the layers of the neural network. For each training case, the switch randomly selectively disables each of the feature detectors in accordance with a preconfigured probability. The weights from each training case are then normalized for applying the neural network to test data.”

The another patent from a development from the same team, however, is based on a spurious claim that they developed the idea for a parallel convolutional network. Hinton and his team have been influential in creating an improved GPU-based implementation of parallel convolutional network, but in no sense did they invent such a network.

Other AI-related patents filed by Google include Q learning with a neural network (invented by Watkins, not Google), classifying data objects (defined so broadly that it could impact all other AI research), and word embeddings. As reddit user AnonMLResearcher said of the latter:

“I am afraid that Google has just started an arms race, which could do significant damage to academic research in machine learning. Now it’s likely that other companies using machine learning will rush to patent every research idea that was developed in part by their employees. We have all been in a prisoner’s dilemma situation, and Google just defected. Now researchers will guard their ideas much more combatively, given that it’s now fair game to patent these ideas, and big money is at stake.”

Google’s artificial intelligence patents appear designed to protect its financial interests, but their specious claims will give them undue credit and hamstring smaller company’s research in the field. And, let’s face it; any company looks small under Google.

Thank you I Programmer for providing us with this information.

Image courtesy of Digital Trends.

Topics: , , , , ,