February 26, 2019 | CogWorld on FORBES
Artificial Intelligence (AI) has been around for about 75 years now. The creation of IBM Watson was a seminal moment that brought AI into the mainstream. I was fortunate enough to be a part of the founding team of Watson and got to experience this paradigm shift firsthand.
Today, AI has grabbed a foothold in enterprise IT and is poised to become as ubiquitous as the internet—touching every aspect of our lives, from our jobs and transportation systems, to financial services, healthcare, digital commerce and entertainment. It's also gaining momentum because of an advancement within AI called augmented intelligence, based on the notion that the real power of AI is not about replacing what humans do, but rather augmenting it. Analysts and industry experts are starting to agree that augmented intelligence is the most effective way to maximize business outcomes and customer experiences.
But while organizations are aware of the benefits they can achieve with augmented intelligence, many still struggle first with where and how to get started, then how to achieve AI success repeatedly and later how to ensure that their AI systems stay within the guardrails of their organization’s ethics. These struggles are answered with AI that is practical, scalable and responsible.
Throughout human history, people have asked questions to get answers. Early AI efforts reflected that same Q&A model. But what if you, as an employee or customer, don’t even know what is possible and therefore what you could and should be asking? Most organizations struggle to find the right data and thus the right business problems for AI. From identifying the business problem that needs the most attention, to providing expert insights for a specific industry, to quickly delivering AI systems that deliver the value promised, Practical AI is about accelerating time to market. Today, banks know that applying AI to business process intelligence will revolutionize the way they gather information and interact with customers—and that they can use it to understand the stated and unstated intentions of customers to enable better interactions.
We all want the ability to repeatedly churn out enterprise AI systems, and to consistently manage them. Because of the specialized nature of AI systems, AI needs to become Scalable as an enterprise-wide capability in order for it to succeed. It’s about leveraging people, processes, partners and software to repeatedly deploy and manage enterprise AI at scale. For example, now, using an AI-powered debt risk advisor, a hospital can identify early and prioritize high ‘bad debt’ risk patient accounts. It can, at the same time, use AI-powered patient scheduling to map patient preferences to complex facility, personnel, imaging and equipment resource scheduling, resulting in higher patient satisfaction and better resource utilization.
While AI is becoming more accepted in the enterprise, we have noticed that the conversation is starting to shift from whether AI will work, to questions around fairness, ethics, explainability and compliance of models and underlying data ecosystems that power these AI systems. Chatbots, recommendation engines and virtual assistants are found in our offices, our kitchens and even our pockets.
Given the potential power of AI—and the free rein humans may be willing to allow it, parameters need to be set. There is an urgent need for building AI systems that are free from bias, transparent in their operations and can reflect the core values and policies of the business. Banks, cancer centers, defense and intel organizations are wary of ‘black boxes’ that give random answers, even if the answer is right. Trust is something that builds with time, and trust can be accelerated by clearly communicating rationale. The need for evidence-based insights becomes that much more critical when dealing with heavily regulated industries.
The pertinent question is what are we all doing to ensure that our future with machines is a bright one? For AI to positively impact individuals and communities, it needs to take a ‘people and ethics first’ approach—it needs to be Responsible.
As AI continues to evolve, it has the potential to transform the world at a scale larger and more profound than the industrial and agricultural revolutions. But, miles to go before we sleep! We all need to collaboratively and aggressively realize the great promise of AI while mitigating its perils, all towards a greater good.
Manoj Saxena, columnist, likes building things, going fast, and helping brilliant people build great companies. Manoj also likes making markets especially around machine intelligence. He is passionate about racing cars and inspiring our youth to pursue STEM careers.
Currently a venture capitalist, Manoj is focused on cognitive computing and machine intelligence, and the Saxena Family Foundation. He is a former CEO and founder of two successful venture funded software start ups which were acquired by IBM (2006) and by CommerceOne (2001). Manoj is also Founding General Partner of The Entrepreneurs' Fund, a $100M seed fund. He led the IBM Watson Software Division as its first General Manager (2011-14) and helped with the formation of the Watson Business Group in January 2014 with a $1B investment from IBM. Manoj holds nine software patents in machine intelligence.