How Artificial Intelligence Can Help Enterprises Make Data More Productive

Data is the asset that drives value in the technologically-driven knowledge economy. And yet, for all its emphasis on efficiency and productivity, enterprises waste a lot of that value because they don't know how to identify what's important and how to interpret it.

The real benefit of newer technologies like artificial intelligence may be in the decidedly unglamorous effort to optimize its meaning and impact. JL

Anne Gherini reports in Inc:

Over $3 trillion is lost each year due to bad data. Extracting data is not the challenge anymore. (It) lies in making that data accurate and actionable to the end user. If your customers can't swiftly turn that data into business results (without engineer or analyst support), the data is limited in value.
The field of Artificial Intelligence (AI) was first introduced at a conference in 1956. At the time, AI founder, Herbert Simon, predicted, "machines will be capable, within twenty years, of doing any work a man can do." But what Simon and the other optimistic early AI experts failed to appreciate are the numerous challenges involved in relying solely on machines - entities that lack the emotion that informs the human decision making process.
Despite the fact that the AI field was born more than half a century ago, we've only recently in the last few years made monumental strides in advancing the artificial intelligence game. More specifically, we've started to transition from a model whereby AI assists humans to one whereby humans assist AI. While the difference sounds subtle, the shift empowers us to advance one step closer towards creating more efficient and thriving businesses.
The Power of a Bot
The near future of brand communication will undoubtedly entail some element of messaging bots. Messaging apps surpassed social media apps in daily active users (DAU) in 2016. Marketers are still warming up to the idea of leveraging these platforms to access the data generated by their vast user bases.
One of the most prominent use cases for chatbots is content distribution. The inbound movement, coupled with the death of banner ads, has led marketers to embrace content creation as an effective means through which to engage with customers. The challenge is it is now crowded. Given that, distribution is key to content marketing success, bots are quickly proving to be an attractive instrument to engage and communicate with a desired audience.
A second rapidly growing use case for chatbots is customer service. AI-assisted bots can seamlessly handle and process mounds of frequently asked questions in real time without any human input. Companies like, Amex and Macy's have been utilizing bots to scale customer service quickly and keep costs down.
A recent study from Accenture found that one of the ripest opportunities for bots relates to the fact that managers devote the bulk of their time to coordination and control tasks. This "routine work" is a prime candidate for bot assistance. Companies like Conversica have been leveraging conversational AI to accelerate the sales cycle by automating the outreach, engagement and qualification of sales leads, thereby freeing up the human salespeople to spend more time in real conversations with their prospects.
Data as a Natural Resource In addition to spearheading the birth of the more consumer-facing chatbots, human-assisted AI has also taken big steps towards helping business solve some of our most pressing big data problems. Over $3 trillion is lost each year due to bad data. Extracting data is not the challenge anymore. Now, the challenge lies in making that data accurate and actionable to the end user. "Data is like a natural resource, you have to cultivate it to extract value," says IBM's Caitlin Halferty.
Several companies are leveraging their expertise in data science to make big data more actionable for the average business. At Node.io, we use the web as our primary data source. The we rely on natural language processing and machine learning to vet and make that data actionable to sales and marketing. The data layer is a fundamental element, but if your customers can't swiftly turn that data into business results (without engineer or analyst support), the data is limited in value.
The Output is Only as Good as Your Input
While it is an exciting time seeing the future of business in an AI driven world, it is also conceivable and quite probable that the current and future rate at which AI advances might surpass the rate at which businesses adopt this intelligence. The reason is that adoption requires adaptation. Businesses need to not only test these new products and services, but also dedicate team members to help teach these "machines". As Accenture found, in order to be effective, machines must be trained in contexts of every given situation - the "learning" in machine learning - and that is not a fleeting task