Services that use some form of artificial intelligence to augment an end-user experience are now well on their way to mainstream adoption. The latest example of this manifested itself this week at the Microsoft Build 2016 conference in the form of “chatbots” that Microsoft plans to make available as digital assistants that can be infused into almost any application. The goal is to make it possible to invoke a chatbot as a service and to provide a framework through which organizations can create their own chatbots.
That move comes on the heels of a more ambitious IBM Watson initiative under which IBM has exposed more than 30 application programming interfaces (APIs) to developers. The initiative has already lead to the development of several relatively simple applications, such as The North Face using Watson services to help online shoppers find products and Hilton building a concierge robot that promises to transform the hotel guest experience.
Other major AI initiatives include everything from the DeepMind Project underway at Google to Apple’s recent acquisition of a company that uses AI to read facial expressions of users as they sit in front of a screen.
Obviously, it’s still early days when it comes to AI. In fact, there’s no better example of what can go wrong than Microsoft’s recent misadventure on Twitter with Tay chatbot, which was hijacked by a few mean-spirited end users who taught it to curse and make racist and misogynist comments. As embarrassing as that may have been for Microsoft, it’s another example of how accessible AI technology is becoming.
AI and the IT service provider
Arguably, all these instances of AI are now coming to the forefront because of the rise of the cloud. There is now enough inexpensive compute power available to drive advancements in AI, and there’s no doubt that the availability of all that compute power will lead to further AI enhancements.
In the meantime, all these advancements present a unique opportunity for IT services providers. In its simplest form, AI initially shows up in most IT environments in the form of machine learning algorithms that automate a specific function. From there, it’s a short leap to more cognitive computing applications that use AI to augment a task performed by humans. In fact, the IBM Watson platform is already being used by managed service providers to augment the help desk experience.
For all intents and purposes, the number of use cases for AI is only limited by the imagination. The opportunity for IT service providers now is starting those “what if” conversations with customers. It still takes a while to train an AI service to understand a particular body of knowledge, but once they have a core understanding of the task, these systems quickly start to train themselves. Naturally, not everyone is comfortable with that idea. But, like it or not it is happening. At this point, the only remaining question isn’t if AI will impact an organization, but when.