IBM fiddles with Apple on new AI services

Two AI offerings

At the IBM Think conference here, IBM disclosed two offerings: IBM Watson Services for Core ML and the IBM Cloud Developer Console for Apple.

IBM Watson Services for Core ML lets developers build enterprise-grade AI services and apps that deliver real-time insights and address user queries, according to Angel Diaz, IBM's vice president of developer advocacy and technology. And the applications get smarter with every new user interaction.

Apple describes Core ML as a foundational machine learning framework used across Apple products, such as Siri, Camera, and QuickType. It is optimized for on-device performance to minimize memory footprint and power consumption.

When combined with IBM''s Watson cloud-based services, Core ML helps developers build secure AI services and apps that connect to their enterprise data.

IBM provides Watson services, such as Watson Visual Recognition, and enables developers to bring in Core ML models and run them on the IBM Cloud.

"The use cases you''ll start to see will be around image recognition and things like field force automation," Diaz said. For instance, a field technician could take a picture of an unfamiliar machine part and have it identified by an application built from this capability, he said.

Developers can train their model locally and then update the master model on the IBM Cloud, Diaz said. "So the model gets smarter; we''re democratizing the training of the models themselves," he said.

 'Apple, through its acquisitions, WWDC presentations and other announcements, has demonstrated that it has the machine learning and AI "religion," said Sam Charrington, principal analyst at CloudPulse Strategies, St. Louis. "The problem is they're really weak when it comes to cloud, and cloud services are an extremely important delivery model for ML and AI," he said. "Extending their partnership with IBM to incorporate IBM Watson services helps them fill this critical gap, and is a win for developers using Swift and Core ML."

The other piece, the IBM Cloud Developer Console for Apple, provides automatic configuration of a developer''s development environment, integration with a DevOps toolchain, and easy access to add AI services and applications.

The developer console enables Swift developers to easily link to the IBM Cloud to build apps and comes with step-by-step guidance for developers of various experience levels, along with integration with AI, data, and mobile services optimized for the Swift programming language.

Also, the console comes with the IBM Cloud Hyper Protect Starter Kit, which enables iOS developers to safeguard credentials, services and data that use the IBM Cloud Hyper Protect services.

The fact that Apple has to tap IBM's cloud-based Watson capabilities to provide a deep level of machine learning and AI capabilities, is an admission that IBM Watson's AI capabilities fill key gaps in the Apple product line, said Ronald Schmelzer, senior analyst at Cognilytica, an analytics consulting firm in Ellicott City, MD.

"Likewise, the fact that IBM has struck a partnership with Apple to let Apple iOS mobile app developers leverage IBM Watson cloud capabilities is testament to the fact that IBM needs more developers to leverage their platform, and Apple commands a huge army of app developers that can instantly be turned on to the IBM Watson platform," Schmelzer said.

Learning to FfDL

Over time, these Apple developers that move to adopt the IBM Cloud might also use Watson Studio – IBM''s service to build machine learning workflows and train and deploy machine learning models at scale.

Watson Studio features the new Deep Learning as a Service capability, which is based on IBM''s Fabric for Deep Learning (FfDL, pronounced "fiddle") technology, which the company has open-sourced.

"Just as The Linux Foundation worked with IBM, Google, Red Hat and others to establish the open governance community for Kubernetes with the Cloud Native Computing Foundation, we see IBM''s release of Fabric for Deep Learning, or FfDL, as an opportunity to work with the open source community to align related open source projects, taking one more step toward making deep learning accessible," said Jim Zemlin, executive director of the Linux Foundation, in a statement. "We think its origin as an IBM product will appeal to open source developers and enterprise end users."

So, you might say that IBM is teaching Apple developers how to FfDL.