Provided by

Augmenting people, processes, and potential

December 17, 2019
AI technologies promise to extend human capabilities, writes SAS chief executive Jim Goodnight, and ultimately, improve the world around us.

From mass surveillance to mind-reading machines, each new day seems to bring another alarming prediction about the potential of artificial intelligence to change the world.

But we don’t give enough attention to the practical AI applications that are in use every day. These real-world applications aren’t creepy or futuristic. You might even call some of them mundane. But they provide practical value to businesses and consumers, and they aren’t leading us to impending doom.

AI does have the potential to change our world. But it’s not going to do that through sentient robots or computers that take control away from humans. The AI applications that we see will more often augment human activity than replace it.

As we continue to witness increased computing power and a more connected world, practical AI technologies like natural language processingcomputer vision, and especially machine learning will proliferate and become even more useful. These are the practical applications of AI that will improve our lives.

Data scientists review computer vision heat maps that are used in identifying animal tracks for conservation efforts. Courtesy of SAS

Natural language processing. Today you have access to AI right in your pocket. I can ask my cell phone what the temperature is outside or what time the grocery store down the road will close for the day. My phone understands my question, accesses the answer online, and then answers my question in my language.

The ability to interpret and process language is a fast-growing use for AI. Outside of personal uses on our phones, natural language generation can provide context for the results you see in a report, explaining the output of a machine learning model or asking whether you want further analyses. These verbal explanations can make complex analytics accessible to more people, even if they don’t have a deep background in analytics.

Natural language processing also powers chatbots that provide information to customers online. If you go to sas.com and choose to chat, we’re using our own technology to engage with customers. We still have live, human chat representatives. But for many of the requests people have, a computer can provide automated responses. That frees up our chat staff to focus on more complex questions.

Computer vision. Another way to demonstrate AI is through computer vision. This method continues to find new and innovative use cases.

Doctors are using computer vision to measure and classify tumors on medical images. Conservationists are using computer vision to analyze photos of animal footprints to help monitor endangered species without invasive tags.

Of course, computer vision is also crucial for self-driving vehicles, including cars, trucks, and commercial vehicles used in plants and warehouses. Several companies are pioneering vehicles that use computer vision and sensors to “see” the world around them. The technology gives the vehicles 360-degree views, with lasers that detect objects up to hundreds of meters away.

Computer vision is also powering the future of retail, from registerless checkout to visual-style recommendations to predicting demand.

Machine learning. New uses of AI and predictive analytics have the potential to change almost every facet of every organization. Take anti-money laundering (AML) technology. Hundreds of organizations worldwide use AML to help identify problematic or illegal financial transactions across the globe. By adding AI and machine learning to existing AML technology, we’ve seen a reduction of false positives by 50% to 70%. That means fewer transactions for manual intervention, leaving staff to focus more on cases that are truly harmful to the company and its customers.

New uses of AI and predictive analytics have the potential to change almost every facet of every organization.

Many of my favorite AI use cases have this same element of making humans more productive. AI and machine learning can improve and augment important work that’s already being done around the world. They take what already works and makes it better.

Many people ask which machine learning models should be used for which problems. Let’s review four types of machine learning models and describe some scenarios in which they are commonly used.

  1. Neural networks. In the financial world, neural networks are helping investigators find and stop fraud by uncovering trends across millions of transactions. ERGO, a German insurance company, is using predictive analytics from SAS to find unjustified claims. Customers also use neural networks in their buildings to optimize power usage and predict mechanical failures.
  2. Decision trees. Through the use of decision trees, we’ve helped rapidly growing Wake County, North Carolina, make property tax assessments fairer and more accurate. We are working with many tax authorities to uncover tax fraud and find citizens that have underdeclared their income.
  3. Regression methods. Regression methods are a tried-and-true statistical practice, and they are finding new life in the AI era. We can use this technology to better understand target markets. For government agencies, regression methods can help identify fraud and waste in existing programs. For banks and financial services companies, it can lead to better risk assessments and create the foundation for an adaptive, more accurate anti-money laundering program.
  4. Forecasting. Forecasting allows large government programs to account for changes in populations and understand how these shifts affect government services. In the electric generation industry, companies are constantly monitoring the grid and examining weather data to make sure the power is there when you need it. Eni—an Italian energy company—uses SAS predictive analytics to control, clean, and prepare geologic well data. This helps Eni automate and refine its oil exploration process.

Many factors are driving the hype and interest in AI: the internet, the digitization of analog data, the increased use of images and video to communicate, the internet of things, and our ability to capture and store this data with cloud technologies.

After more than 40 years of developing and implementing analytics software, I have an optimistic view of technology and its capability to augment and amplify our human potential.

While AI will disrupt and change certain jobs and industries, we as humans have the curiosity and passion to direct these powerful technologies to achieve what was once impossible. Creativity, problem solving, and the ability to collaborate with diverse teams will be even more important as we move into the future with AI. Together with technology, humans will continue to improve our lives and the world around us.

Jim Goodnight is co-founder and CEO at SAS.