Does your business need Artificial Intelligence (AI) or Ambient Intelligence?
In the past 4 years I have worked with countless Analytics, AI and Business Intelligence (BI) vendors that guaranteed that their services, products and or solutions would deliver real-time data for every customer from which I could derive insights, process and decisions. As a Technology Executive for the past 12 years and an Enterprise Architect for the previous 10, I have had the privilege to explore, conceptualize, develop and implement technological advancement that presented an in-depth overview of the customer lifecycle experience. Most of those experiences centered around data capture, aggregation, correlation and analytics to derive actionable insights. Over time, speed and scale took precedent over analysis as Cloud computing accelerated the pace of data acquisition. With an extensive experience in hospitality, gaming and travel, I always thought that I could use technology, data and analytics to build the perfect framework to encapsulate a customer journey.
“Unfortunately, I realized early on that you must include a contextual awareness factor into any equations that involves humans interaction to meet an expected outcome.”
Even though AI will ultimately deliver on the promise of ubiquitous computing, I believe that Technology vendors have not resolved the most obvious problem related to the presentation of viable and relatable insights. It all comes down to context, you can acquire customer data across different Omni-channels in the like of voice, sms, email, applications, chats, database, meta-data etc… But how do you validate in which context such engagement took place. Did the customer react to an unexpected reply? Did the business fail to deliver on the product or service? You must also take into account the bias that programmers (guilty as charged, I’m one of them) may inadvertently introduce into the data acquisition process.
One the most used or misused platform for insights nowadays relates to surveys. Most surveys occurs after the fact, based on a set of conditions for which rarely capture the essence and the context of the engagement. As myself I responded to countless surveys over the years and not once did a customer support representative contacted me and addressed my concerns based on comments that I had left. In most cases and when it really mattered to me I had to “Personally” drive the engagement and often had to reestablish the context so the representative could “fully” understand the narrative and resolve the issue. Fittingly many businesses and vendors alike continue to invest resources and I would say have improved on their overall customer interactions and have started to build some awareness into their solutions to track challenges and or respond to an engagement.
We still live in a world where AI targets queries and algorithms, we constantly hear about algorithms that can predict the future, inherently this may work if you use constant variables like weather and traffic data to validate an outcome. How do you predict that Sam, your most sought after concierge will fall hill and negatively impact the experience of several of your Platinum customers as only Sam posses the contextual interaction with the customer.
I believe that one may want to consider Ambient Computing and Ambient Intelligence as prerequisite to Artificial Intelligence. As such Ambient Computing encompass the combination of hardware, software, user experience and machine learning. The Internet of Thing (IoT) lends itself as one of the preamble behind an improved Contextual Awareness.
You may want to think of Ambient Intelligence as a Neural Network for which would help support data awareness and work to enhance people day-to-day activities in a simple and intuitive way based on the analysis provided.
As such, you would conceptualize that an application would “sense” and “understand” the context and then respond to the human factor as a data point and build the Contextual Awareness to derive a set of “Prescriptive” outcomes. I prefer the term Prescriptive vs Predictive even though I do see a value in a prediction, but like most I believe that the key to meet an expected outcome relates to one ability to prescribe the action required to meet that outcome. Then we can finally engage and talk about ubiquitous computing and Artificial Intelligence.
Eric Saint-Marc, you can follow me on twitter @thinkahead