Separating AI Reality from AI Hype

That means that it will not take over the world but rather some tasks, functions, and even some jobs (much as I hate to say that) that have been historically manual endeavors. But that’s also its value.

AI: THE ADVANTAGES

When it comes to business, AI is becoming increasingly important as a means of achieving these distinct benefits:

1. It represents the next evolution in business process automation. Called autonomics and/or robotic process automation, this application of AI involves using it and machine learning to automate complex tasks that are governed by rules but that typically need human involvement. Deloitte, in its paper “Intelligent Automation Rules the World,” cites the example of Genworth Financial, which uses AI to automate long-term-care underwriting. The firm uses a combination of fuzzy logic, an evolutionary algorithm, and natural language to come up with the business rules and processes needed to figure out what is best for a particular customer and meets all compliance requirements. What’s interesting is that it has been doing this since 2004!!

2. It can help attain higher degrees of personalization. There is no doubt that customers are demanding highly individualized interactions with companies, the sense that the company (a) “knows” them, and (b) doesn’t waste their time with offers, opportunities, interactions, and experiences they don’t want and can’t control. One great example of this is subscription-based Stitch Fix, whose personal clothing and accessory stylists make individualized selections for female and male subscribers. The stylists work with 60 data scientists to tailor clothing and accessories to members’ tastes, lifestyles, and, of course, budgets. An initial questionnaire sets a baseline, asking questions such as blousy or close-fitting tops? Favorite colors? Or my personal favorite: “Are you more urban hipster, Sex in the City chick, or Bohemian?” Once the baseline is set, the stylists, both human and AI, go to work selecting the clothes and accessories and sending each member a personalized selection. The member keeps what she wants and returns the rest. The machines learn from the selections, and the next batch is even closer to the personalized customer preferences. The results are amazing. More than 80 percent of clients return within 90 days for a second order, and a third of clients spend 50 percent of their clothing budget with the subscription service.

3. It can advance baseline sales, marketing, and customer support functionality. The basic capabilities these pillars of CRM have long provided will get a significant upgrade. A good example is Conversica, which is focusing on using AI for lead generation, or Cogito, which uses AI, machine learning, and advanced algorithms to detect the emotional state of a customer calling in to the contact center and the customer service rep’s response to the customer’s state. Cogito’s technology then provides personalized suggestions on how to either advance or recoup the conversation.

There are dozens to hundreds of other uses of AI that have proven value. This is one trend that won’t be a trend much longer, because it will simply be part of the next evolution of mainstream technology—perhaps not to the extent of mobile technology, but the idea is the same. It will just become part of the way things are done.

That said, we’re still in the early stages. Forecasters peg the AI/machine learning market at somewhere between $12 billion and $16 billion by 2020. There are outliers like Tractica, which predicts $36 billion by 2025, but overall growth is expected to be more conservative.

But what makes AI important and powerful is that its use cases keep expanding. Even with these conservative projections, the tech giants are all jumping into the fray—Salesforce (via its platform entry, Einstein), Oracle (via its AI add-on for its enterprise solutions), Microsoft (which has interwoven AI into its applications framework), and SAP (which calls it “Application Intelligence” and follows a similar path to Oracle, embedding it into its enterprise solutions) are all heavily involved in the fray, as are hundreds of niche players.

What we have established is that AI’s value is real, the uses cases are innumerable, the hype is scary but fictitious, and its evolution is still in the early stages. Is it worthy of your company’s time and investment? I don’t know. Maybe my algorithms do.

---

Paul Greenberg is the managing principal of The 56 Group, a customer strategy company. He is the author of CRM at the Speed of Light, which is in nine languages and is currently in its fourth edition. He is also the author of the upcoming Commonwealth of Self Interest, to be published next year.