…We can easily prove that Information, Knowledge, and Wisdom can all be represented as Data. However, no one can prove that Data, Information, Knowledge, and Wisdom (DIKW) are all different things. Maybe, it’s time to accept that everything is just Data.
Ackoff’s DIKW Theory
If you’re a Knowledge Management theorist or practitioner, you’ve surely come across Russell Ackoff’s theoretical notion of the “Data, Information, Knowledge, Wisdom” (DIKW) hierarchy. (Ackoff, 1989). However, there’s a major problem with Ackoff’s theory… After many decades, there has been no evidence to prove its accuracy or value. So, if you believe in it, you may be missing the most important notion of KM, which is that Information, Knowledge and Wisdom are all just Data.
Engineers have proven that, to date, absolutely nothing can be communicated from any one person or systems to any another person or system without ensuring that what is being communicated is first transformed to the least common transmittable denominator, which is “data.”
For humans, data is communicated in many forms that conform with our human senses…
- Our ears can sense audible frequencies that represent audio-data.
- Our eyes take in photons that travel over physical frequencies, which represent visual-data.
- Our skin and fingers can sense physical surface and temperature changes that represent touch-related data.
- Our noses can detect particles in the air that represent smell-related data.
- Our tongues can sense variations in particles and temperatures for taste-related data.
For machines, data is communicated in the form of electronic and mechanical impulses, waves, and movements…
- Strumming or picking a guitar’s strings generates audible frequencies for humans or electronic frequencies for electronic amplifiers.
- Computer screens create arranged and lit pixels at different visual frequencies that can be picked up by eyes or light detectors.
- Mechanical systems can automatically create Brail messages that human fingers or machine sensors can detect.
- Mechanical systems can mix chemical combinations that mimic the scents detected by the human nose (think “scratch-and-sniff”).
- Mechanical systems can mix chemicals to create artificial tastes that mirror many of those in nature.
In short, anything that can be communicated between a person and/or a system to any other person and/or a system must be communicated in the form of data.
So, what is Information?
Ackoff states with no supporting facts that “Information is contained in descriptions, answers to questions that begin with such words as who, what, when, where, and how many.” However, if you know the answer to any of those questions, you actually “know.”
The truth is that there is no provable distinction between Data and Information. In fact, it can be successfully debated that one person’s Data is another person’s Information. For example, if we look at the symbol “1”, a computer scientist or a mathematician might view it as a discrete value. However, to an engineer, it can also represent an electrical signal or impulse that is conveyed in either a digital or analog form and that is transformed, through electronic circuitry, to represent a set of ordered and colored pixels which have been grouped together to create the human or machine detectable symbol “1”. So, is “1” Data or is “1” Information? The answer lies in the truth… “1” is both Data and Information, lending credence to the belief that “Information = Data“.
So, what is Knowledge?
Ackoff makes the statement that “Knowledge is conveyed by instructions, answers to how-to questions.”
However, it can be easily proven that how to do something can be scripted in the form of algorithms that can be temporarily or permanently persisted, such as work instructions, recipes, and computer code, such that a machine can follow the details of the algorithm, automatically. So, if you know how to do something and you want to convey what you know to another person or to a machine, how can you possible convey it without representing it as some form of transmittable data? The proof is in the fact that you can persist what you know. And, in order to persist what you know, it must be in Data form. In other words, we can strongly make the case that “Knowledge = Data” because you cannot possibly convey what you know to anyone or anything else without first converting it to Data. And, once persisted, the machine that holds that persistence now knows what you know (or knew). So, it can be debated that the machine knows what you know. Just ask Google or Amazon Alexa a question and you’ll find that some machines know far more than most humans know about many things.
So, what is Wisdom?
According to Ackoff, “wisdom is the ability to increase effectiveness.”
Years ago, we would say things like “Granny or Grandpa are full of wisdom” because they had a way of communicating high probability outcomes that could often be taken as unarguable truths. These outcomes had been tried and tested many times, over many years, but Granny and Grandpa seemed to be able to communicate those truths in simple short sentences. However, it was not until we personally tested those statements, ourselves, that we could be sure of the outputs and then say something like “I should have listed to them, from the beginning.” The more we tested Granny’s and Grandpa’s statements, and the more their statements proved to be true, the more we learned that they were right and full of wisdom.
If you thought that only humans could do this, welcome to the world of Machine Learning (ML). Computer Scientists can now program computers to learn from outcomes and to even improve their data and algorithms. So, it can now be argued that machines have wisdom because they can learn through algorithms and data to improve their outcomes and to prove that certain actions or calculations will lead to much higher probability outcomes than other actions and outcomes. In summary, “Wisdom = Data“.
A Practical Example: Development of Knowledge Repositories
One very practical area where we see the benefits of treating (at least) Information and Knowledge is in the area of Knowledge Repository (KRs) generation. In the past, KM professionals used traditional Content Management Systems (CMSs) to manually author, format, categorize, organize, interlink and publish knowledge that was deemed important for sharing, in long narrative formats, such as we see in Wikis like Wikipedia. However, unlike Wikipedia which has many millions of editors, most enterprises have small groups of people who have very difficult times authoring, publishing and maintaining large volumes of content. In fact, a common complaint is that most traditionally developed KRs quickly become stale and unmaintainable. For this reason, more and more enterprise are moving to automated generation of their KRs, directly from data, which exists in abundance in most enterprises. This paradigm is called Data Driven Synthesis (DDS) of Knowledge Artifacts and Knowledge Repositories and uses automation to significantly outperform human output.
The original version of the KMBOK was generated in a Wiki that required manual authoring, formatting, classifying, organizing, inter-linking and publishing of content. It took multiple people many months to develop and, over time, a great deal of its contents became stale with many dead links. In short, it became more stale and more difficult to maintain, over time. As a result, its use by end users started to die off.
The new version of the KMBOK leverages data-driven automation. It treats all Information and Knowledge as nothing more than Data and assumes that from Data, everything else humans need can be automatically generated. In other words, the tool that generated it automatically turns “data” into human consumable “web content”. As a result, the new KMBOK was developed in about two (2) hours. Maintenance of the data-generated KR, which can easily include the regeneration of many thousands of pages, formats, and links, rarely takes more than just a few minutes. And, because its generation was founded on automatically converting data into content, it has orders of magnitude more content pages, views, knowledge constructs, and live HTML links. It also has far more advanced visualization features. In short, the newly generated KMBOK was easier to create, is easier to maintain, and offers more to its end users, all because it treats Information and Knowledge as nothing more than Data.
Conclusion: Everything is Data
So, all historical facts accumulated in the last few decades of technology further fortify the notion that DIKW is not a progression, such as bottom to top hierarchy of “D -> I -> K -> W” but is more of an equality, such as “D = I = K = W”, because nothing there is nothing we can think of or communicate that is not, first, transformed to data. And, if it can exist as data and be reconstituted as something more than data by a receiving person or system, it is still just data, even after reconstitution.
If you believe this is not true, simple show the counter facts. To date, no one has been able to successfully do so.
Impact to you as a KM Professional
Here’s what you need to know as a Knowledge Management professional. Accepting that D = I = K = W ensures that you don’t waste time trying to address and come up with solutions that are inconsistent for each of the four elements in the equation. The Information Technology (IT) industry has proven, over the last 4 to 6 decades, that computing machines can recreate human functions such as but not limited to: Search, Indexing, Data Mining, Natural Language Processing, Machine Learning, Business Intelligence, Data Visualizations, Automated Inference (Engines), Voice Recognition, and much more. On the other hand, over a much longer period of existence the non-technical Knowledge Management (KM) industry has not even come up with one major and transformative achievement to its name.
So, if you want to get to more efficient and effective enterprise knowledge solutions, don’t waste time on “IKW” because the IT industry has been successfully proving that treating everything as “D” will get you and your stakeholders much farther, much faster.
- Semantic Wikis – Pros and Cons
- Intranet Management Best Practices and Guidelines
- Using Enterprise Knowledge Maps to Achieve Better Enterprise Knowledge Management
- Ackoff, R. L. (1989). From Data to Wisdom. Journal of Applied Systems Analysis, Vol. 16, p 3-9.