Building a bigger toolbox for AI developers

For developers, more tools are better—and open source tools are often the most powerful of all.

There are more than 20 million software developers worldwide, and a growing percentage are focused on advancing tomorrow’s essential technology—applications enabled by machine learning and artificial intelligence. What do they all need more of, to speed their ideas? Tools.

For developers, more tools are better—and open source tools are often the most powerful of all. Open source developer platforms are essentially massive workshops where developers can create, collaborate, learn new tricks and put fresh spins on old ones. And of all the dynamic technologies revolutionizing the enterprise—cloud computing, IoT, blockchain—no technology is sparking developers’ ingenuity more than machine learning and AI.

That’s one reason why IBM is giving AI developers a much bigger toolbox. On March 20, the company is making a new set of AI and machine learning tools and resources available to developers everywhere. As Angel Diaz, IBM’s Vice President of Open Source and Developer Advocacy, explains, “With this announcement, we’re making it easier for developers to build AI models, to train those models and then to rally as a community to drive the art of the possible in open source.”

Here are four of IBM’s new AI tools and open-source platforms aimed at empowering the work of developers worldwide:

IBM and Apple unite for AI. Building on a three-year partnership, IBM and Apple are combining enterprise AI and machine learning technologies in a new joint offering called Watson Services for Core ML. They are also introducing a new cloud developer console with step-by-step guidance for developers of all experience levels that will make it easier and faster to build apps. Before, it could take days to code an app; now developers will be able to do it in a matter of minutes.

Dive into deep learning. Another first: IBM is offering Watson Studio’s Deep Learning Service as an open-source project, free to all developers. The Fabric for Deep Learning or FfDL (pronounced “fiddle”) is a scalable, deep learning platform aimed at helping developers create and train deep learning models more quickly. It can be used on IBM Cloud with a range of popular open-source frameworks, such as TensorFlow, Caffe and PyTorch, lowering the barrier to entry for any developer to get started with coding AI applications.

Sharing to the MAX. IBM is also opening up its new Model Asset eXchange (MAX), an online forum where developers can put up their models for others to use, evaluate and share ideas.  Nothing is for sale; every model on MAX is free and meant to inspire other developers.

A community center for AI. Two years ago, IBM launched its Spark Technology Center to spread the adoption of Apache Spark technology. Now, the center’s mission has expanded into AI. Renamed the Center of Open Data and AI Technologies (CoDAIT), it’s a forum where AI developers can collaborate with each other and a team of experts from across the IBM spectrum, including IBM Watson, IBM Research and DGB. CoDAIT’s ambitious goal: to bring open-source practitioners together to make AI models easier to create, deploy and manage across the enterprise.