Google Cloud Platform has launched two machine learning APIs into open beta. The software tools are part of a larger initiative by the company to use its artificial intelligence (AI) tools to help customers improve businesses and build new business models.
In addition, Google made its Tensorflow machine language software APIs available. Tensorflow lets developers build machine learning models and then scale them to production. Google uses Tensorflow internally to build machine learning into some of its consumer products.
Google’s parent, Alphabet, is pitching these AI software APIs to commercial customers as a way to better compete with public clouds such as Microsoft Azure and Amazon Web Services (AWS). Both Azure and Amazon have more than twice Google‘s amount of revenues from cloud computing.
Google is a big proponent of AI and has used the technology developed by its DeepMind subsidiary to make its data centers more efficient. For example, a business could use the tools to transcribe phone calls or help automate certain parts of phone conversations.
Google also announced that customers can now rent storage, computing power and other cloud services from its data center in Oregon. The company plans to add a dozen new data centers during the next 12 to 18 months.
Google’s AI push isn’t particularly surprising. Earlier this year during the company’s Google Cloud Platform conference, GCP NEXT, Alphabet Chairman Eric Schmidt touted the benefits of machine learning and said that he believed the combination of cloud, crowd sourced information, and machine learning would be the basis for every successful IPO win in the next five years.
But Google isn’t the only tech company exploring the benefits of AI. Earlier this week, Macy’s Department Stores debuted an in-store, AI-powered assistant that uses IBM’s Watson. The app lets shoppers input natural language questions and get customized responses.
Plus, Microsoft also touts an Azure Machine Learning Studio, where customers can experiment with various tools and analytics and eventually deploy the service.