IBM is seeking to increase the use of artificial intelligence across enterprises. This is through three different strands. First, is via a new ‘Deep Learning as a Service’ initiative. This lets users run hundreds of deep learning training models at once and quickly optimize their neural networks. With this service, users only need to pay per graphics processing unit. The aim here is to turn something that was once a cost-prohibitive capability into somatic more accessible to businesses.
IBM’s Watson is a question answering computer system which is capable of answering questions posed in natural language. The system came out of IBM’s DeepQA project. Over time Watson has become cloud based and it has evolved machine learning capabilities, with optimized hardware available to developers and researchers. The announcement was made at the recent IBM THINK conference, aimed at showcasing business applications.
The second tranche is that IBM has open-sourced the core of its ‘Deep Learning as a Service’ offering. This is termed ‘Fabric for Deep Learning’ (or fiddle for short). This move embraces a wide array of popular open source frameworks and also lowers the barrier to entry for deep learning. This is a cloud-enabled offering.
The third area is with the unveiling of MAX (Model Asset Xchange) and CODAIT (the Center for Open-Source Data & AI Technologies), which are set to assist businesses further. With these new tools, TechCrunch reports, developers can develop their models with the same open source frameworks they are likely already using (think TensorFlow, Caffe, PyTorch, Keras and so on).
At the center of each of these initiatives is the concept of ‘Deep Learning as a Service.’ This is intended to make extracting data easier, overcoming the common barriers to deep learning deployment, which are seen as skills, standardization, and complexity. According to IBM: “By helping to streamline and accelerate the development process for data scientists and AI engineers, companies can now quickly extract rich insights, create more engaging consumer experiences and ultimately drive greater business value.”
In related news, Google is also paving the way for artificial intelligence and machine learning, and the company has launched a new service designed to enable both businesses to begin building their own artificial intelligence systems. The system is called the Google Cloud AutoML.
