Harnessing the Data Explosion with Artificial Intelligence
Every second, 2.6 billion emails are sent. Each day brings 400,000 variations of malware. Each year, 80 million MRIs (magnetic resonance imaging scans) are performed. In one year, more than 100,000 articles are written about cancer, and 70,000 are written about cybersecurity.
The amount of data being generated around the world is exploding, as illustrated by Beth Smith, general manager of technology of IBM Watson. Speaking at the 2018 KeyBanc Capital Markets 20th Annual Technology Leadership Forum, held in Vail, Colorado, Smith underscored the need for businesses to not just manage data, but also to learn from it—and how deep learning artificial intelligence (AI) can help.
The Opportunity: Using AI to Turn Data into Knowledge
As Smith explained, advances in machine learning, deep learning, cloud and core technology make it possible to turn data into knowledge. Only 20 percent of data is publicly available, while the other 80 percent is inside interfaces, behind firewalls, in the middle of workflows and business processes, and in supply chains. By tapping into that 80 percent, industries can unlock decision support capabilities that are estimated to grow to $2 trillion by 2025.
Looking at the banking industry for example: many are using chatbots to interact with consumers, making customer service timelier, improving satisfaction, and reducing the need for agents in the moment. Smith notes that, by 2020, chatbot and whatever the next-iteration of conversation technology becomes will account for $8 billion of value for banks around the world.
How IBM Watson is Changing Business
Smith leads the research and development team for IBM Watson, a suite of enterprise-ready AI services, applications and tools based on the IBM cloud that learns from private, shared, industry and public data to deliver solutions with industry specializations. Smith explained that IBM Watson has three key focuses:
- Learn more with less data: Software based on machine learning typically needs to analyze a significant volume of data before it provides accurate insights or predictions. Enterprises house an enormous amount of data, but most of it is not properly formatted and governed for machine learning systems. Companies need machine learning applications that can learn more and produce accurate results with less training data.
- Create better workflows: IBM Watson can be embedded within business workflows, so it can deliver AI solutions when and where the enterprise needs it.
- Protect customers’ insights. IBM doesn’t treat data as a commodity, but rather as a valued asset for its customers. IBM Watson enables customers to protect their data and extra value out of their data models, while retaining ownership of their data.
IBM has also addressed the growing concern around trust in AI with its Principles of Trust and Transparency:
- The purpose of AI is to augment human intelligence.
- Data and insights belong to their creator.
- AI systems must be transparent and explainable.
How Companies Are Using AI and Machine Learning
Smith offered examples of how industries are putting machine learning systems to work. For instance, The Royal Bank of Scotland (RBS) uses Watson to support Cora, a digital customer support avatar that handles 35 percent of customer inquiries. Cora can successfully address 85 percent of the calls, allowing RBS’ human support staff to focus on more complex issues.
At Sogeti, the technology and engineering services division of Capgemini, Watson is being used as a cyber-detective. Watson helps security analysts in Sogeti’s cybersecurity division mine vast amounts of information to identify threats and their causes, reducing identification time down to minutes instead of hours.
A Look at the Future of AI
In just the last three decades, digital technology has rapidly evolved from “computer in your pocket” smartphones to advancements in networks and connected systems that have enabled the massive growth of Google and Facebook. According to Smith, the next era of technological advancement is going to focus on knowledge and how it can be used with emerging technologies such as block chain.
She describes the evolution from “narrow AI,” or the idea that AI can handle a particular task, to “broad AI,” in which an application takes the learning from one task and applies it to a very different task. For example, an insurance company could use a broad AI application to evaluate images of any kind of home damage, whether a hurricane, tornado or other disaster.
Another shift has happened in the way in which technology is delivered. In the past, a company would need a highly customized deployment, usually with the help of consultants, for its data and applications. Today, companies can use a self-service platform, such as a cloud-based application programming interface (API), to develop software programs or applications to use their data.
AI: The Key to Understanding and Using Data
Data is transforming every industry. Deep learning AI can analyze large volumes of data at lightning speed to help organizations make faster, better decisions. As we learned at our 20th Annual Technology Leadership Forum, a new marketplace is taking shape around helping businesses mine their data to accelerate internal workflows, improve processes, develop customer solutions, and create new products and services.
- Data is growing exponentially, and 80 percent of it is inside business interfaces and not available to the public.
- Deep learning artificial intelligence (AI) can help turn data into knowledge.
- IBM Watson is an enterprise-ready AI service that helps businesses draw more value from their data.
- AI is being used by many industries to improve processes, handle customer service, and analyze data quickly to deliver solutions.
- The future of AI lies in interconnecting with other emerging technologies and applying learning from one task to a different task.
To learn more about our Annual Technology Leadership Forum and how we’re preparing companies with insights to drive growth contact Arvind Ramnani at 503-821-3897 or email@example.com.