Issue link: https://read.uberflip.com/i/1444470
Document #US43529118TM ©2020 IDC. www.idc.com | Page 4 IDC White Paper | Accelerate Machine Learning Development to Build Intelligent Applications Faster Other use cases include office applications surfacing related content and suggestions as knowledge workers develop new analysis reports or create new content for a project. Situation Overview Introduction The market for deep learning-based artificial intelligence applications has grown rapidly and continues to surge. IDC estimates that spending on AI applications will exceed $63 billion by 2021, growing to over $96 billion by 2023, and by 2025, at least 90% of new enterprise application releases will include embedded AI-based functionality, recommendations, or advice. Organizations need to consider the following reasons as to why these systems are important for their future: » Augment human judgment. The best business cases are about extending human capabilities, not replacing them, by positioning intelligent applications as an extension of human intention. Power tools in the hands of a craftsperson is the best analogy. Pricing optimization models are good examples of deep learning in this area. A second example would be an AI imaging application that automatically identifies cancerous tumors by examining radiology images, aiding radiologists. » Accelerate investigation and discovery. Even the very best human readers cannot comprehend millions of pages of documents in one day. Applications that understand natural language can be applied to this task for both the spoken and the printed word. Deep learning-based natural language systems provide better results than handcrafted, taxonomy-based systems. » Recommend "next best actions" and predict outcomes. Deep learning- based applications build models using relevant data for recommendations and predictions, which are some of the typical use cases. » Personalize outcomes and recommendations. Many organizations are beginning to use deep learning models to "personalize" content, predictions, and recommendations to specific customers or prospects. This is especially true with mobile applications where users increasingly expect their devices and applications to "know" their likes, dislikes, and expectations. » Automate organizational knowledge management. While knowledge management systems have existed for decades, many have failed under the weight of human effort required for ongoing operation. Applying automation to investigation and discovery activities, or developing best practices, is a key benefit. Automatic categorization and theme identification of documents are some of the key use cases of deep learning. Other use cases include office applications surfacing related content and suggestions as knowledge workers develop new analysis reports or create new content for a project.