I like to think of artificial intelligence as a metaphor for human intelligence and function: seeing, thinking, acting, and learning. Seeing is about pattern detection and recognition; it is about our senses that we have available to us. Thinking is where we form plans, weigh options, make predictions, and optimize against a variety of constraints and objectives. Acting is where our thoughts and plans, generated from experience and perceptions, are manifested to cause interactions and outcomes with other people and the environment. Learning takes near and long term feedback from the environment to improve our internal algorithms for seeing, thinking, acting, and learning.
Over the past decade, we have seen tremendous achievements in making Sensing more available (e.g. Internet of Things, advances in machine learning), Thinking more powerful (e.g. Big Data, deep learning), and Acting more achievable (think robotics but also the societal transformations where ubiquitous computing, mobile and wearables are becoming part of our society’s fabric). No one of these achievements has been a tipping point of artificial intelligence. Rather, the combinations of achievements in seeing, thinking, and acting have opened a wealth of new opportunities that artificial intelligence is poised to address. This momentous time for the field is what prompted me a few weeks ago to move from upstate NY to Seattle to join the Maana team.
After completing my university degrees, I was lucky to join one of the oldest corporate R&D labs in the world and develop technical and leadership skills over a wide range of real-world industrial problems. A couple of key lessons I learned: identifying the customer need and a path to tangible value is a critical driver of innovation, teams are vital for enjoying work and success, and it is possible to do things in an agile way (I recently gave a talk on adapting the agile methodology to R&D). I also learned that big innovations require taking calculated risks, and opportunities to do so do not always come around twice. Identifying these magic moments is difficult. One of the requirements of R&D is to think about what’s coming next, what is possible, and what will be valuable. My intuition told me that the artificial intelligence hype was not just hype, but that the combination of multiple technical, cultural, and business factors was creating an opportunity. One that would create value-adding technology that assists humans to do meaningful work while increasing productivity, enable new types of work, and allow people to interact with other people in creative and meaningful ways—avoiding mundane and repetitive tasks performed in isolation. After exploring many different places to pursue this opportunity, I landed in Seattle at an AI startup called Maana.
Maana is a Knowledge Platform company that delivers both a process of approaching the above opportunity as well as a technology to make it both possible and efficient to build knowledge-driven applications in an iterative and agile way. First, we work with business executives to identify their objectives. We then enable subject-matter experts in the organization to model and decompose those objectives into questions that answer them. Lastly, our platform allows end users to rapidly connect those questions with underlying data and analytics. This contrasts with the Big Data era that promoted a type of bottom-up, data-first approach, which led to the current demand for data scientists to search for value within large corporate data systems. The Big Data approach often is inefficient and frustrating for everyone involved as the data-capable are not always knowledge-able about the business objectives. At Maana, we’ve learned that our customers have the best knowledge of what questions are meaningful. By putting the knowledge workers at the center of the Maana Knowledge Platform™ and encouraging them to start with those questions—not with data capture, processing, and analytics—we can get to value faster and more efficiently.
Therefore, we’ve created a two-fold capability: a novel process that defines the key question and answer structure with our customer in the form of knowledge models, and a powerful distributed Maana Knowledge Graph that sits at the intersection of AI technologies and low-latency compute environments to rapidly index, integrate, and query enterprise data. These two capabilities work in concert: the Maana Knowledge Graph is the network of models that provide daily recommendations to assist subject-matter experts in making better decisions at much faster speeds, , and the Maana Knowledge Models act only as the definition of key business questions and the structure of the answer connected to real data, but also as an archive and documentation for the organization to capture their corporate knowledge for collaboration and persistence. By tying these two entities together—the definition of questions and answers, with the actual compute mechanism—we believe we can overcome the loss of knowledge that many enterprises suffer from as well as create a powerful environment for the reuse and innovation of the Knowledge Models.
I’m excited about my role at Maana as the Chief Scientist. I will be driving our customer-focused and applied research organization and agenda, and creating the next wave of features that excite our customers and deliver value. As you can imagine, there are several opportunities where research can provide value to creating, improving, and managing knowledge models; as well as assist in performing valuable data processing and analytic tasks. But most importantly, I’m excited to see what the Maana team can do together with our customers to bring new value and capabilities to society.
Stay in the know with the latest information about Maana services, events, news and best practices by email.