Getting to the future of KM
For all the talk of AI moving to center stage in knowledge management, much less is said about the true impact of these new technologies on knowledge workers themselves. The fact of the matter is that automation will have a considerable impact on KM over the coming years, some of it bad for knowledge workers, but some possibly good. To understand how those changes and impacts may look, we first need to grasp the current state of technology in KM.
The current state of KM
Put simply, the current state is typically a complex mess, and while AI tools can significantly improve the situation, no amount of AI, robotic process automation, or blockchain technology will fix it completely. We have multiple disconnected silos of information, mountains of duplicate files, poor search functionality, and often little-to-no structure or governance applied to knowledge or information assets. AI can read and index vast amounts of disparate information with a high degree of accuracy, and it can dramatically improve the quality, accuracy, and experience of search.
Still, there are many things AI cannot do—and many more that it can do well, but arguably not well enough. Here’s the thing: AI can handle the sheer scale of information volumes with ease and in a manner that no team of human knowledge management workers could, but that does not mean it does so with 100% accuracy. And that is what can cause the wheels to come off the AI/KM bus, depending on what you are trying to achieve.
Knowledge work automation
In order to understand why, let’s be clear on what knowledge work automation is. McKinsey defines it as “the use of computers to perform tasks that rely on complex analyses, subtle judgments, and creative problem-solving.” I think that is a solid definition that illustrates where automation projects can run into trouble. In practical terms, “to perform tasks that rely on complex analyses” is where automation projects excel. The legal profession has been embracing AI for some time now to trawl through mountains of information and pull together briefs or evidence. Intelligence and law enforcement have been using AI successfully to spot trends, patterns, and bad actors. In theory, this is a good thing, and when AI assists human experts in their work and helps with and speeds up some of the heavy lifting, most people are happy. But the technology industry sees the role of humans in these situations as temporary, not permanent. This is how AI technology vendors think of it:
Stage 1: AI assists human knowledge workers—helping them do what they already do better, faster, and more accurately.
Stage 2: AI augments human knowledge workers—enabling them to do things they couldn’t have done previously.
Stage 3: AI conducts autonomous knowledge work—removing the need for a human knowledge worker.
That is how a lot of AI operates and learns, regardless of its focus. It first assists you, then learns from what you do, starts to suggest better working methods, and finally takes on the work for itself.