The recent announcement of Salesforce Einstein -- dubbed "artificial intelligence for everyone" -- sheds new light on the new and pervasive usage of artificial intelligence in every aspect of businesses.
From consumer to business
Even without knowing it, the consumer world has long been using AI-based systems. Siri knows which movies are showing at what time and recommends the best theater based on the user's location. Nest learns about household habits to optimize heating and cooling patterns. Facebook recognizes friends in photos with nearly 98 percent accuracy.
The beauty of these systems is that they surface AI-driven functions, without ever exposing the user to the complexity of artificial intelligence models, deep learning or machine learning. Data scientists aren't optional -- they are completely out of the picture. At least on the surface.
The move toward AIaaS
In a blog post announcing Einstein, Jim Sinai, VP of Marketing for SalesforceIQ, provided a summary of the power of this new AI system -- and the value for the company:
Powered by advanced machine learning, deep learning, predictive analytics, natural language processing and smart data discovery, Einstein's models will be automatically customized for every single customer, and it will learn, self-tune, and get smarter with every interaction and additional piece of data. Most importantly, Einstein's intelligence will be embedded within the context of business, automatically discovering relevant insights, predicting future behavior, proactively recommending best next actions and even automating tasks.
Just as Salesforce is no longer just Software as a Service CRM, but also a complete Platform as a Service for many types of apps to run, let's venture a guess that Einstein will at some point be made available "as a Service" for other Force.com and Heroku apps to use. Here comes a new breed of Platform as a Service: AIaaS (Artificial Intelligence as a Service).
Chatbots, another breed of AIaaS
Chatbots, or conversational bots, are the "other" trending topic in the field of artificial intelligence. At the juncture of consumer and business, they provide the ability for an AI-based system to interact with users through a headless interface. It does not matter whether a messaging app is used, or a speech-to-text system, or even another app -- the chatbot is front-end agnostic.
Since the user does not have the ability to provide context around the discussion, he just asks questions in natural language to an AI-driven backend that is tasked with figuring this context and looking for the right answer.
Chatbots embody the essence of AI-powered backends, made available through APIs, for any authorized frontend to consume. Through the use of RESTful services, their integration into any app, web site, or any other frontend is made easy.
Unlike what Einstein could become however, chatbots are purpose-built applications. They represent a different type of Artificial Intelligence as a Service, closer to Software as a Service than to Platform as a Service.