A lot of the buzz around artificial intelligence (A.I.) and machine learning has focused on self-driving cars and voice-activated digital assistants—very flashy, consumer-centric applications. But A.I. might end up having its largest impact on the enterprise space, where companies need such platforms to analyze and act on massive datasets. Take Salesforce, for example, which has tried to integrate A.I. features onto its CRM offerings via Salesforce Einstein, its A.I.-services platform. Earlier this month, the company unveiled Einstein Sentiment, which judges user sentiment (positive, negative, neutral) in any sort of text (emails, texts, and so on); Einstein Intent, which tries to determine the intent of customer inquiries in order to route those customers to the appropriate customer-service or sales representative; and Einstein Object Detection, which identifies objects within images. Taken together, these three tools can amplify a company’s ability to handle customers. Public perception of a firm might gradually improve, for example, if customer-service reps can use an A.I. to proactively offer updated shipping information or refunds before they even pick up the phone and talk to an irate shopper. (In terms of object detection, there are probably firms that need that sort of specialized functionality—Google has developed something similar, clearly in an attempt to service a developer need.) In some ways, the A.I. utilized by Salesforce Einstein is a step beyond what Facebook, Microsoft, and other firms are offering with bots, which can automate big chunks of the customer-service experience via scripts. But bots—especially at this relatively early stage in their evolution—can’t really read customer nuances or sentiment. Earlier this year, research firm Gartner estimated that, within a year, 80 percent of data scientists’ toolboxes could include deep learning or deep neural networks. That’s perhaps a bold prediction, considering how many businesses are only just beginning to tiptoe toward these technologies; but the point remains that the enterprise’s relationship with A.I. will only deepen (pun intended) in coming years. For tech pros interested in artificial intelligence and machine learning, and who make their money by developing apps and services for businesses, the route forward seems pretty clear: businesses will begin demanding more sophisticated and nuanced A.I. abilities. Fortunately, it seems that software vendors are already attempting to meet that need with the right tools. Next Big Challenge for Business A.I.: Nuance
A lot of the buzz around artificial intelligence (A.I.) and machine learning has focused on self-driving cars and voice-activated digital assistants—very flashy, consumer-centric applications. But A.I. might end up having its largest impact on the enterprise space, where companies need such platforms to analyze and act on massive datasets. Take Salesforce, for example, which has tried to integrate A.I. features onto its CRM offerings via Salesforce Einstein, its A.I.-services platform. Earlier this month, the company unveiled Einstein Sentiment, which judges user sentiment (positive, negative, neutral) in any sort of text (emails, texts, and so on); Einstein Intent, which tries to determine the intent of customer inquiries in order to route those customers to the appropriate customer-service or sales representative; and Einstein Object Detection, which identifies objects within images. Taken together, these three tools can amplify a company’s ability to handle customers. Public perception of a firm might gradually improve, for example, if customer-service reps can use an A.I. to proactively offer updated shipping information or refunds before they even pick up the phone and talk to an irate shopper. (In terms of object detection, there are probably firms that need that sort of specialized functionality—Google has developed something similar, clearly in an attempt to service a developer need.) In some ways, the A.I. utilized by Salesforce Einstein is a step beyond what Facebook, Microsoft, and other firms are offering with bots, which can automate big chunks of the customer-service experience via scripts. But bots—especially at this relatively early stage in their evolution—can’t really read customer nuances or sentiment. Earlier this year, research firm Gartner estimated that, within a year, 80 percent of data scientists’ toolboxes could include deep learning or deep neural networks. That’s perhaps a bold prediction, considering how many businesses are only just beginning to tiptoe toward these technologies; but the point remains that the enterprise’s relationship with A.I. will only deepen (pun intended) in coming years. For tech pros interested in artificial intelligence and machine learning, and who make their money by developing apps and services for businesses, the route forward seems pretty clear: businesses will begin demanding more sophisticated and nuanced A.I. abilities. Fortunately, it seems that software vendors are already attempting to meet that need with the right tools. 
