Knowledge management requires good information

Knowledge Article Old

Processing data is only the first step in making information truly valuable.

(Note: I wrote this article way back, in 1998, and just found it again. The topic was, at that time, information processing within a web environment, which was pretty new. While reading it anew, I realized most of it could also be applied to A.I., so I just made a few adjustments, highlighted in cursive. The rest is original from almost 30 years ago and still, in my opinion, fully valid. It was written originally in Spanish.)

Information has singular importance. It can produce stock market movements, trigger corporate restructurings, or alter government policies. Furthermore, it has the definitive advantage that, unlike physical products, it can be used an infinite number of times, simultaneously, and in different places.

The Artificial Intelligence (AI) boom is comparable to the creation of information-related companies during the popularization of the web, that significantly surpassed manufacturing activity. Everyone is preparing to take advantage of the competitive advantages, but, although there are many, both then and now, practical experience reveals following five points:

1. Technology alone is not enough

A sophisticated system that allows processing, exchanging, and accessing any type of information via the web can be used for purposes for which it was not designed. We see it every day on social media, with its fake news and the spread of radical messages, or with cryptocurrencies, used for laundering criminal funds. AI is no exception, as it feeds on information available on the web, and its own algorithms can lead to erroneous responses and cause harm to its users.

Given that these tools are firmly intertwined with our workplaces, it seems necessary to incorporate other changes in the usual behavior of workers, educating them on correct, legal, and productive use.

2. Don't collect information

Some organizations that have fully embraced data warehousing and data mining practices, precursors to AI,have an excess of stored data and a procedure for updating that data, but it is very complex to convert it into useful information in a timely manner. Managers need information quickly to act, and if they don't have it, they end up distrusting the system. AI partially solves this problem, as it produces answers in short periods of time. However, it requires very precise prompting to provide the desired results.

A clear rule: "Most information is useless. Don't hesitate to discard it."

3. Information overload leads to indecision

In the quest to share explicit knowledge, some organizations find that after making a significant effort to make it accessible in their corporate, internal, or external tools, they are only utilizing 20% ​​of its potential. When a person encounters many references while searching for information, they tend to use it less or self-limit their options: according to Forbes, only 6% of Google searches lead to visits to the second page of results.

4. The poor reliability or degree of accuracy of the information

If the information is unreliable, it can lead to errors that not only result in incorrect decisions but also incur high business or legal costs. In this regard, two practices are necessary: ​​the continuous maintenance of information and its classification according to clear criteria. It is equally necessary to identify the sources: in the public sphere, a newspaper article does not carry the same weight as a report from an official body, nor, within a company, does an advertising brochure carry the same weight as a technical specification.

5. The bureaucratization of knowledge

If people are to contribute their knowledge, it seems logical to introduce some kind of control to filter out what is not valuable. In practice, institutional figures or curators have been created to validate knowledge, or committees of experts to approve contributions. The result is bureaucracy and arbitrariness surrounding the knowledge incorporation process, leading to the same ineffectiveness and disappointment as suggestion boxes in some organizations.

These are some of the main problems, but, as it is a recent practice, there are surely more to discover. What can be observed from accumulated experience is that, for a knowledge management system to work, it must be accompanied by a change in mindset and habits.

Replacing the idea of ​​punishing mistakes with the idea of ​​experimenting to learn, or introducing the habit of sharing experiences with others and reflecting on them, are just two of the challenges we must face.

Are you willing to do it?