Or AI Made Easy – Two ways companies can monetise AI
It will be many years before the thinking machines of science fiction become reality. No system today can match the reasoning abilities and conversational skills of HAL 9000, the onboard computer from the movie 2001 a Space Odyssey, or mimic humans with the lifelike precision of the androids on HBO’s hit reboot of Westworld. But key aspects of artificial intelligence (AI) are ready for prime time, and they’re taking the world by storm.
AI is powering an ever-expanding universe of smarter gadgets, wearables, homes, cars, and factories. Through enabling technologies like machine learning, natural language processing (NPL), and the Internet of Things (IoT), companies are quickly adding AI functionality to everything from financial services, medicine, education, communications, and customer self-service, says Brendan O’Brien, chief innovation officer and co-founder, Aria Systems.
Two pathways to monetisation
There are two main ways to monetise AI. Indirectly, by making AI part of your products or services, or directly, by selling AI capabilities to customers in use for particular problems or to build their own AI-enhanced offerings.
Indirect monetisation
With indirect monetisation, built-in AI capabilities contribute to an offering’s overall value, but are not the sole source of that value. Take the recommendation engines used by Netflix and Amazon, which make use of advanced machine learning technologies and algorithms. While customers appreciate the recommendations that AI generates, they’re just one of many factors that motivate customers to subscribe to these services by making them more useful and enticing.
So how can you add AI to your offerings? Companies with deep pockets can emulate Google, Apple, Facebook, and telecoms and invest millions hiring teams of engineers, launching skunkworks, and buying up AI startups. The rest of us can consider acquiring AI as a service.
Direct monetisation: AI as a Service
“AI for everyone.” That’s how Salesforce describes its new AI service, Einstein. It captures the essence behind the growing field of AI-as-a-Service (AIaaS). Einstein enables companies to acquire highly sophisticated capabilities with minimal investment, paid for incrementally through subscriptions or usage-based mechanisms.
IBM and its Watson platform are another AIaaS. Watson understands human speech and can find answers to extremely complex problems in seconds. Watson first made headlines back in 2011 when it beat reigning champs at Jeopardy. Since then, it’s helped doctors at Sloan Kettering Cancer Research Centre and the Cleveland Clinic make better decisions about patient treatment and helped H&R Block create virtual assistants smart enough to complete income tax forms.
Amazon Machine Learning is a prime example. According to a recent forecast, revenues from Machine Learning as a Service (MLaaS) are expected to hit $20B a year by 2025.
Another area within AIaaS may be even larger. Natural Language Processing as a Service (NLPaaS) has been around for some time, perhaps best exemplified in the speech to text services provided by industry leader Nuance Communications. Other NLPaaS providers such as Speechamatics and Vocapia now offer similar services.
When it comes to monetisation, the most pervasive use of NLPaaS these days is in chatbots, computer assistants that comprehend text and speech. They’re everywhere. Apple Siri, Microsoft Cortana, Amazon Echo, and Google Assistant are all versions of chatbots.
For example, Facebook Messenger hosts more than 10,000 chatbots – the avatars you engage with in chat sessions on web sites. It’s true that many of today’s phone chatbots are not all that bright. But they’re getting smarter every day and newer chatbots can handle increasingly difficult questions, place orders, escalate services issues, and more—all without human intervention.
The wave of chatbots is only just beginning. Thanks to chatbot-as-a-service platforms like Chatfuel, any company can quickly build their own fully-featured chatbox in just minutes.
The implications of AI
AI for hire is the great leveler when it comes to competition. For minimal cost, you can turn your understaffed workforce into a world-class customer service outfit backed by an army of brainy bots who work 24/7 and never call in sick. Your processes can get more efficient, and your offerings more responsive and personalised.
Companies that can use AI to better understand their customers as individuals, treat them with the care they deserve, and fulfill their ever-changing needs will thrive as the IoT shifts into high gear.
The author of this blog is Brendan O’Brien, chief innovation officer and co-founder, Aria Systems
Comment on this article below or via Twitter: @IoTNow OR @jcIoTnow