Looking at the future of knowledge management in the age of smart machines at KMWorld 2019
At KMWorld 2019, Euan Semple, director, conference chair, and author, Euan Semple Ltd., moderated an opening keynote panel discussion about how AI and machine learning will impact workers and the world, the importance of diversity of thought in training AI, and the need to remove bias from AI and automated processes.
The panel included Anthony J Rhem, CEO/principal consultant, A. J. Rhem & Associates and author, Knowledge Management in Practice; Phaedra Boinodiris, m ember IBM Academy of Technology (Blockchain, Games, Watson, Design), IBM; and Ross Smith, Modern Workplace Supportability Engineering, CSS, Microsoft and Royal Society of Arts, University College Dublin.
Videos of KMWorld 2019 awards presentations, keynotes, and sessions can be found here.
Setting the stage for the discussion, Semple made the observation that humans are uniquely capable of dealing with change, unplanned and unpredictable circumstances, and problem solving.
In fact, ironically, some of the skills that used to make people reliably employable in repetitive, bureaucratic, predictable jobs—being dependable, accurate, not thinking too much, and doing things by rote—are the characteristics that in the future will make them more easily replaceable. On the other hand, being thoughtful, being a troublemaker, and looking at things in different ways and talking about them will make people more employable, he observed.
We, as a people, may be stumbling into some troubling times in the future in fact, if we do not continue to be thoughtful and really consider what we are doing and the consequences and morality of some of what is happening for example in social media platforms, said Semple. This goes back the beginning of knowledge management and considering why and how we do what we do, and its value.
As the capture, delivery, and access of knowledge evolves, AI will contribute to a new and rapidly changing knowledge landscape, according to Rhem. AI plays an important part in KM by elevating how the delivery of knowledge occurs to the people who need it, such as employees, vendors, and customers. It is used to scale the volume and effectiveness of knowledge distribution and delivery through prediction of trending knowledge areas/topics that users need; identification of which targeted knowledge will resonate with audiences based on real-time engagement and content consumption; auto-curation and personalization of knowledge based on individual preferences; and improvement of content decisions by determining precise distribution schedules, paired with machine learning around what content will resonate best with certain audiences.
In terms of smart machines in use today, Rhem said, RPA is providing organizations the ability to automate certain repetitive tasks eliminating the need for additional staff and freeing workers to concentrate on more important task. Chatbots with natural language processing are providing value for employees, vendors, and customers with personalization of the delivery of knowledge, and will provide cognitive capabilities to understand, interpret and manipulate human language that will enable chatbots to anticipate the needs, attitudes and aspirations of users to deliver personalized outcomes.
AI will also make search and its search products more relevant, precise, and efficient, understanding natural language. AI through intents will be able to better know what content your customer needs. Intents will provide a better understanding of what the customer is looking for by better understanding each customer intended use of the content.
However, Rhem said, we all need to pay more attention to how these things are developed who is developing them, and including diversity of thought, people s background and what they bring to the thought processes, and a variety of data. "There is bias in everything we do," he said, noting that it is necessary to train algorithms to remove bias, which, he acknowledged, is almost impossible to accomplish.
Boinodiris cited inspiring uses of AI, such as for disaster preparedness bots to help children prepare to be resilient in the face of disasters. However, she noted, unintentional misuse is also a risk with AI. The intent may be honorable but because people may have inherent bias, what they create can actually do harm, said Boinodiris.
The bias, once it is in AI, can get "calcified" although people may not recognize it, she said. In the future, AI will affect, judicial systems, medicine, finance, security, hiring, said Boinodiris, who gave several examples, including the seemingly innocuous use of an mobile phone application used by drivers that automatically sent data about cars hitting potholes back to a city's department of public works so that they could be repaired. But, she said, the people that downloaded that app onto their smartphones were those who were more affluent, and so those were the people whose streets were being fixed. "We must insist on trustworthy AI," said Boinodiris, who said, "This is all fixable."
Smith considered the historical precedent of technology replacing workers, the car replacing the horse and the ATM replacing the bank teller. Today, every company is a tech company, said Smith, who reflected on what this means for the workplace of the future and how AI will affect humans.
Many speakers at KMWorld 2019 are making their presentations available at www.kmworld.com/Conference/2019/Presentations.aspx.
KMWorld 2019 is a part of a unique program of five co-located conferences, which also includes Enterprise Search & Discovery, Office 365 Symposium, Taxonomy Boot Camp, and Text Analytics Forum.
SAVE THE DATE FOR KMWORLD 2020- NOVEMBER 16-19, 2020!