03 Apr 2020
The ongoing quest of Artificial Intelligence (AI) is for developing it to the level that it matches the capabilities of the human mind, also in certain cases (computations and memory) to even surpass the capabilities of the human mind. Current global debate is regarding the feasibility of achieving such a goal.
“Knowledge by definition is the state of knowing something and thus is always tacit, any knowledge codified in the form of a document/literature to make it explicit thus becomes information. It's the way we assimilate information that leads to the creation of knowledge.”
The industrial age automated mundane manual tasks using machines and left humans to do higher value knowledge work. The knowledge/scientific age aims to take away the burden of knowledge also from humans and leave them with creative work and other intelligences or living on UBI (Universal Basic Income) and performing more entertainment than productive activities.
Knowledge management has provided a foundation for many of the most exciting developments in business today. Artificial Intelligence and Machine Learning including organizational network analysis and industry network development, are proving to be fundamental business tools. In the 21st century, knowledge management seems to reach its next level. Artificial intelligence & Blockchain comes into play and transforms again how knowledge is captured, developed, shared, and efficiently used within organizations.
As the proliferation of structured and unstructured data continues to grow, we will continue to have a need to uncover the knowledge contained within these big data resources.
AI capabilities that can recognize context, concepts, and meaning are opening up exciting new pathways for collaboration between knowledge workers and machines.
Experts can now provide more of their own insights for training, quality control, and fine-tuning of AI outcomes. Machines can amplify human prowess and create new experts. This will greatly affect knowledge workers all over the world. To fully use AI to their advantage, companies will have to redesign knowledge worker’s flows, experience and roles.
The Challenges For KM Today
The earlier KM implementations couldn’t serve the purpose as they were highly oriented towards converting tacit to explicit and then making the explicit discoverable and reusable through repositories and forums powered by search.
There are strong criticisms about the possibility to ‘convert’ tacit into explicit knowledge (Nonaka and Takeuichi, SECI model for knowledge creation, 1995). The biggest failures of any attempt to manage knowledge is the challenge of “tacit-explicit” conversions.
Tacit knowledge is articulable, in-articulable or somewhere in between. Workplace skills are articulable, and sense-making, meaning, wisdom, emotions and feelings inarticulable (Busch, 2008). Moreover, it is either very difficult or impossible to codify it in true sense.
This puts added burden of externalisation on people who are either not motivated, disciplined or capable enough to do it. All such codification efforts either result in poor articulation thus poor assimilation by consumers of knowledge or failure to keep the codified information up to date.
Knowledge workers are people who reason, create, decide, and apply insights in non-routine cognitive processes.
More than 150 such experts drawn from a larger global survey on AI in the enterprise, almost 60% say their old job descriptions are rapidly becoming obsolete in light of their new collaborations with AI.
Some 70% say they will need training and reskilling (and on-the-job-learning) due to the new requirements for working with AI.
And 85% agree that C-suite executives must get involved in the overall effort of redesigning knowledge work roles and processes.
AI is bringing a great deal of change into the way people work using software but the challenge remains to make software more relevant for users in terms of flows, experience and roles.
In an organization with 1000 employees, approximately $2.7M are lost every year in productivity, according to an independent research done by Panopto.com. 60% of employees report that it is very difficult, or nearly impossible to obtain information from their colleagues needed to do their job. Employees spend 5.3 hours per week waiting for information, sometimes longer. These delays have a major impact on project schedules — 66% will last up to a week, and 12% a month or more. 81% of employees are frustrated when they cannot access the information they need to properly do their job.
Therefore, they might opt to forage information on their own. It can be extremely overwhelming and sloppy, if there is no one to guide them through the process. Every week, an employee spends over 8 hours just doing this.
Also, maybe because of failure to communicate or just unawareness, an employee might be replicating the work that is already done. On average, employees reported spending nearly 6 hours each week duplicating other people’s work.
This portrays the knowledge safeguarding experience is exceptionally valuable, and that turnover negatively impacts the company’s knowledge resources — costing the company both time and money.
The SECI Model Of Knowledge Conversion
Lets revisit the SECI model of knowledge conversion (by Nonaka & Takeuchi 1996) before getting into how artificial intelligence can help achieve it better than the traditional KM approaches.
SECI model of knowledge conversion is the cornerstone of knowledge creation and transfer theory, even today.
They proposed four ways knowledge is shared and created in the organization.
Socialization (tacit to tacit conversion): Knowledge is passed on through conversations, practice, guidance, imitation, and observation. The typical informal sharing of knowledge between people, commonly referred to as “Water Cooler” talks.
Externalization (tacit to explicit conversion): Difficult but important conversion mechanism. Tacit knowledge is codified into documents, manuals, etc. so that it can spread more easily through the organization. Since tacit knowledge is very difficult to codify, if not impossible, the extent of this knowledge conversion mechanism is debatable. Involves writing out documents to share knowledge about projects, products, processes and decisions.
Internalization (explicit to tacit conversion): As explicit sources are used and learned, the knowledge is internalized, modifying the user's existing tacit knowledge.
Combination (explicit to explicit conversion): This is the simplest form. Codified knowledge sources (e.g. documents) are combined to create new knowledge(actually information). It's actually explicit-tacit-explicit conversion as the user is processing the knowledge and applying their intelligence & wisdom to create a new piece of codified knowledge.
Explicit and tacit knowledge blends are such that it’s not possible to use one without the other (Ray and Clegg, 2007; Cohendet, 2014).
Where Can We Get The Maximum Impact?
The table below maps the various skills/attitudes required for effective knowledge conversion across the various types.
As you may have observed that the skills required for each conversion are not uniform across a workforce thus it becomes a challenge to manage it well.
In the last few decades, we seem to have done well on socialization and combination front using search, content authoring and messaging technologies. But it's not enough today as these technologies ended up being (not) adopted differently and in most cases far from being impactful, given the exponential rise in information, produced globally and expected to be consumed to do our work. While we could not effectively get the signals across to people we did manage to increase the noise in our systems.
Liebowitz describes the concept of knowledge management (KM) bluntly as “creating value from an organization’s intangible asset”. The main problem with his concept is to recognize the intangible assets, to address them, turn them into reality and make them accessible for the organization as a whole.
The most impactful conversions are the tacit-explicit (externalisation) and explicit - tacit(internalisation) but the same are the most difficult to achieve due to the various barriers.
The cause of the failures, besides organizational and technological barriers has been the individual behavioral aspects. These aspects in simple terms are
The average human:
- By nature is lazy and not motivated enough to spend extra time to share
- Wishes to get knowledge but not share it
- Views sharing knowledge as a threat
- Is not confident of their communication skills
- Lacks empathy and is driven by selfish motives
- Is like a horse, you can take to water but can't make them drink
How Artificial Intelligence Can Power Knowledge Management Sustainably?
The irony today is that we are being less productive and more distracted given the over use of digital tools to get simple things done. We all suffer from different levels of attention span and flow issues and related chaos/stress.
For better discovery and application of knowledge in an organization or community we need to rework the flows and make them intelligent and assisted. We need to have IALF (Intelligent Assistive Lean Flows).
Smart technologies might close the gap between codification and collaboration. The biggest areas where AI can pitch in is to help organizations in simplifying their workflows by CLADing them for knowledge and collaboration. Provide workflows and digital experience that seamlessly covers the following aspects of managing knowledge in the respective flows:
AI also helps to overcome the past problems of dealing with huge amounts of data, which were deemed unwieldy and difficult to maintain. Modern systems using AI are able to handle big data also providing a certain degree of security using new ways of data storing such as HDFS, NoSQL, decentralized blockchain data storage.
Potential KM Application Areas for AI
The applications need to provide intelligence, assistance or automation of various steps of a lean collaborative workflows that are designed closer to the natural flows of people.
Some areas where real pains can be solved are:
Conversational & Proactive Assistant
Covers aspects of search, productivity, performance (analytics), help, and adaptive learning & discovery. AI provides the means to process human input such as handwriting and voice recognition with the help of the advent of natural language processing. It can help in identifying most relevant content & people(expertise) as well as obsolete content.
In Context Ambient Knowledge
When we perform tasks AI can help by showing data from the past that can be reused or other data points/artefacts that are relevant to the current user context. Idea is to provide all related knowledge to the user at each step of a workflow when they are in the flow. It's about embedding ambient knowledge in the flow to enable better decision making, faster completion and to avoid reinventing the wheel.
AI Microservices For Recommendations
Expert and AI recommender systems help to boost knowledge management and provide the intelligence to perform tasks better. AI can look into your projects and suggest better ways to plan, can look into your leads funnel and provide recent intelligence on prospects, can look into your support funnel and provide insights into SLA compliance or avoid tickets with automated responses.
AI Microservices For Automation
Knowledge management takes advantage of AI tools used to capture, filter, represent or apply knowledge. Using for example knowledge repositories like corporate Wikis , document storages, open data, APIs, AI tools provide applications for the selection, parsing, analysis, classification and even writing of text, automated reasoning and visualizations to facilitate decision-making.
Hewlett Packard: AI can rapidly turn beginners into pros. Hewlett Packard demonstrated that when they used their AI lab’s cognitive computing platform to analyze two years’ worth of call data for a client’s call center. The call center was using a queue-based system for routing customer calls, resulting in long wait times and poor-quality customer support. The cognitive computing platform was able to determine each agent’s unique “micro-skills”—the agent’s knowledge of a specific kind of customer request, captured from previous calls. These micro-skills are now used to match incoming calls to agents who have successfully processed similar requests. The customer support center has seen a 40 percent improvement in first contact resolution and a 50 percent reduction in the rate of transferred calls.
Society of Petroleum Engineers (SPE): The increasing volume of information and diversity of channels make it difficult to connect with the knowledge and subject matter experts needed to solve problems. As a result, cross-industry knowledge flow is impaired. To address this challenge, SPE has implemented a new research portal, supported by artificial intelligence (AI). The portal integrates subject matter expert knowledge with AI natural language processing and machine learning. It automatically enriches documents by classifying them into relevant taxonomies, geo-tagging oil fields, and extracting key concepts, authors, and institutions. These enrichments enable SPE members to zero in on the relevant information from all SPE channels and to graphically analyze timeframes, geography, related concepts, and cross industry collaboration (using social network analysis).
AI tools are maturing rapidly and are due to get a quantum leap in the coming decade. Even efforts to give machines free will, emotions and consciousness are underway.
Engineers and pioneers across disciplines are designing AI so that it is more easily trained and evaluated by experts and can incorporate their extremely valuable and often scarce knowledge.
To begin to take advantage of these new possibilities, organizations will have to allocate their AI spend accordingly. And to get the greatest value out of both their systems and their knowledge workers they will need to reimagine the way specialists and machines interact. Just as today’s machine learning systems augment the capabilities of ordinary workers, tomorrow’s systems will elevate the performance of knowledge workers to previously unattainable levels of uniform excellence.
The world deserves fresh attempts from enterprise software vendors to raise the bar. Time is ripe for enterprise software to tap intelligence to make workflows relevant, leaner, easy, and high RoI for customers.
Strategy, process centric approaches, interorganizational aspects of decision support, research on new technology and academic endeavors in this space will continue to provide insights on how we process big data to enhance decision making and productivity.
To initiate your journey towards higher productivity, try Crrux today, the leanest digital stack for business.
Hirlak, Bengü & Yeşil, Salih. (2019). Exploring Knowledge-Sharing Barriers and Their Implications. 10.4018/978-1-5225-5427-1.ch006.
Sanzogni, Louis & Guzman, Gustavo & Busch, Peter. (2017). Artificial intelligence and knowledge management: questioning the tacit dimension. Prometheus. 35. 1-20. 10.1080/08109028.2017.1364547.