Sinclair Schuller October 3, 2023
“Those who cannot remember the past are condemned to repeat it.” – George Santayana
If you’re a business leader, chances are that you’re not just familiar with this quote – you’ve probably made it a central tenet of your strategy. And for good reason: navigating future situations based on past experience and knowledge has the dual advantage of averting unforced errors and establishing proven patterns for success. The challenge, though, is in the “remembering.” Businesses, like humans, have an imperfect memory.
Business knowledge tends to be tacit. It exists in the minds of individual team members, or even as an ephemeral, collective unconscious that is the organization itself. Trying to capture it is a sisyphean task; documented information is piecemeal and subject to bias, while even perfectly-captured knowledge lacks the wisdom and human reasoning that makes it valuable.
But let’s put the complexities to the side and focus on the potential instead. What if you had near-perfect historical knowledge and the rationale of your predecessors available on-demand? What if you could ask them what they would do given the current context and situation? I think of this as “knowledge permanence”: a concept that was unthinkable just a few months ago, but thanks to the breakthrough of generative AI (GenAI), one that’s poised to become the foundation of organizational success. Let’s explore this idea more deeply.
A Crucial Milestone
A vital cognitive milestone in every human’s life is “object permanence“: the intuitive revelation – one that occurs early in childhood – that objects exist even when they’re out of sight. Businesses can achieve a similar cognitive shift with “knowledge permanence” – the persistent accessibility, relevance, and utility of tacit organizational knowledge.
At its core, knowledge permanence is the ability to maintain and utilize a robust reservoir of knowledge that remains constant, even in the face of staff turnover, technological shifts, and market disruptions. It goes beyond mere data storage; it’s about capturing the holistic context around that data. It’s about permanently capturing the questions and actions that generated the data, the downstream decisions that the data acted as input to, and the immeasurable ripple effect of that data. Making that knowledge actionable and intergenerationally transmissible within an organization is what knowledge permanence is all about.
But knowledge permanence is much more than just collecting and disseminating information. By capturing past experiences and data, it also offers the rare opportunity to break free from outdated mindsets and outgrown assumptions. Rather than falling back on “we do it this way, because that’s how we’ve always done it,” when implemented correctly, generative AI systems use the full spectrum of data to create fresh insights and innovative interpretations based on what’s worked (and what hasn’t). That’s what really makes knowledge permanence so powerful.
Failure to achieve this milestone places organizations at risk of “organizational amnesia,” where valuable insights, protocols, and strategies become lost, redundant, or outdated.
AI Makes Knowledge Permanence Possible
Understanding the value of knowledge permanence is just the first step. The next step (and for builders like me, arguably the more interesting one) is figuring out how to achieve it. Earlier, I posited that it’s only possible with GenAI. Here’s what I mean.
Imagine having a centralized generative AI system in your organization (or even a domain-aligned set of them). The purpose of this system is to provide universal access to (1) collective reasoning, (2) knowledge task execution, and (3) an ever-expanding body of knowledge that continuously elevates the quality and fidelity of (1) & (2). The growth of the core knowledge base is fed from two sources: external information resources, and, crucially, self referentiality. Self referentiality, in this context, is the monitoring and continuous capture of human prompting on (1) and (2), as well as the outputs resulting from those prompts. This is the highest-quality knowledge, and it is the very foundation of knowledge permanence.
So what does it take to make knowledge permanence a reality in your organization? You’ll need to take four broad-stroke actions:
- Establish a centralized generative AI system (or a small set of them).
Standing up systems that will last decades is possible – just ask IBM about their mainframe business.
- Empower and encourage every team member to embrace the system. Building new habits is hard, but it’s easier when that habit demonstrably improves productivity and quality of life.
- Capture all pre-existing knowledge. “Priming” the centralized GenAI system with knowledge is important to provide a “launch zone.” Luckily, companies that provide business systems are making this easier than ever.
- Ensure continuity of knowledge capture. Ensure that the system captures all of the input & output of team members’ usage, and wire every critical external system into the centralized AI to continuously build the knowledge corpus.
How you fill in the details of the coarse-grained outline is up to you, but this framework offers the structure you need to get started. The sooner you begin the journey to knowledge permanence, the sooner you’ll be able to unlock tremendous tactical support by improving speed to value and quality of value on tasks ranging from the repetitive and mundane to the most sophisticated knowledge-oriented work.
At Nuvalence, we’ve started taking steps to achieve knowledge permanence, and are actively guiding client strategy to do the same. One thing is certain: 10 years from now, I wouldn’t want to be part of an organization with amnesia. Would you?