Knowledge velocity is the fundamental driver of change today - and human connection is at its heart

The rapid change we're all experiencing at the moment comes down to one fundamental driver — knowledge is moving faster. This is not just about AI, although that is the most explosive accelerant today. But its impact is built upon advances such as the evolution of the Internet, mobile computing, and all the other ways in which digital technology has been forging new connections over the past few decades — a trend towards what I've been calling Frictionless Enterprise, as previous barriers to information and action get swept away. All of these changes are making it progressively easier than ever before to discover, disseminate, discuss and develop knowledge.

Here's a concrete example from the world of business that for me illuminated the practical impact of this easier access to information. A few months ago I was in a meeting with the CEO of an enterprise technology vendor setting out its growth strategy. He presented stats about its market penetration across all of its verticals and geographies, down to individual countries and sub-sectors. The strategy sought to build on its strengths in each of those markets, with different verticals emphasized according to the specific market profiles and penetration in each country.

Why, I wondered, had such a detailed analysis and point-by-point strategy never been carried out before? While this was a new CEO coming in with a very data-driven approach, this wasn’t new data. It had always existed. But what struck me in that moment was that it had never previously been feasible or economic to collect, analyze and track it in such detail. It is only because of recent advances in data integration, analytics and the support provided by AI tools that this kind of detailed, in-the-moment analysis has become routinely possible. And it’s not just this isolated example. It is becoming progressively easier to analyze more data in more detail than ever before across many different fields. As a result, decision-makers no longer have to guess — which is what they’ve always done in the past. There are still judgments to be made, but they can be based on more knowledge — because it’s moving faster.

Mining new seams of knowledge

The effect is magnified because there are still huge reserves of knowledge that were previously uneconomic to collect and organize. Too often, enterprises look at technology as a means of lowering the cost of what they already do. But this is completely the wrong reaction in the face of rising knowledge velocity. Lowering the cost of handling the knowledge you have already accumulated may have some value, but it is far outweighed by the opportunity to be found in harnessing knowledge that was previously uneconomic to get at. I'm frequently seeing people citing the Jevons Paradox at the moment, and this is absolutely on point. It is named after 19th-century economist William Stanley Jevons, who observed that, as technology increased the efficiency of coal consumption, demand for coal didn't decline — it increased dramatically because so many new uses suddenly became economically viable. In a recent essay, Aaron Levie, CEO of Box, applies the Jevons Paradox to the domain of knowledge:

Jevons paradox is coming to knowledge work. By making it far cheaper to take on any type of task that we can possibly imagine, we’re ultimately going to be doing far more. The vast majority of AI tokens in the future will be used on things we don't even do today as workers: they will be used on the software projects that wouldn't have been started, the contracts that wouldn't have been reviewed, the medical research that wouldn't have been discovered, and the marketing campaign that wouldn't have been launched otherwise.

Even more intriguing than these neglected tasks that we never started before, there are many other hitherto unrecognized seams of knowledge yet to be mined. I recently wrote about the decision traces identified by Foundation Capital's Ashu Garg and Jaya Gupta, part of the often unrecorded context graph that surrounds how an organization makes decisions. Enterprise systems have typically been blind to this often tribal knowledge because it's held in people'e heads and informal exchanges, rather than formally recorded. Exploring a related theme, my colleague Ian Thomas recently highlighted the messy, ad-hoc data that people working at the operational 'knowledge face' typically work with, far away from the tidy data models that populate enterprise applications. He argues that this is the data that enterprises should be focusing their AI initiatives on, rather than applying it to the more limited subset of data they already manage within their existing core enterprise systems.

Human networks of knowledge

The other dimension that's too often neglected when discussing the impact of technologies like AI is how humans adapt to it. I've chosen to use the term knowledge velocity precisely because what matters far more than rapid point-to-point knowledge access by individuals — or by autonomous agents — is the speed at which knowledge can be networked and augmented across human teams and populations.

Look back at any of the technologies in human history that eased access to knowledge, and you'll see parallel rises of institutions and social structures that have accelerated its velocity of circulation. My go-to example is the proliferation of coffee houses across Europe in the 17th and 18th centuries, which became lively meeting places for merchants, scientists, writers and philosophers to discuss the ideas then circulating in newspapers, pamphlets and letters. These amphitheaters for the amplification and recombination of those ideas acted as a seeding ground for the creation of more formal institutions that have become part of the fabric of the modern world — the insurance center of Lloyds of London being the iconic example. Steven Johnson, writing in Where Good Ideas Come From, describes such crucibles of social conversation as 'liquid networks':

When you work alone in an office, peering into a microscope, your ideas can get trapped in place, stuck in your own biases. The social flow of the group conversation turns that private solid state into a liquid network...

It's not that the network itself is smart; it's that the individuals get smarter because they're connected to the network.

Today, the instant global connections of the WorldWide Web have made the entire world a 24x7x365 coffee house for the exchange and dissemination of knowledge and ideas. The vast landscape of open source software has emerged out of this collaborative network, leveraging the pooled knowledge and experience of the online developer community. Collective endeavor has replaced the encyclopedias of yesteryear with the crowdsourced Wikipedia, while the pagerank algorithm that powers Google search is itself based on the collective judgements of humans as they link their own online works to those of others. Today, the conversation moves on across sites like diginomica, on Substack and Medium, disseminated via social networks such as LinkedIn, X and elsewhere — and of course the foundation models of generative AI such as ChatGPT, Claude and Gemini build on all of this collective knowledge, albeit often without acknowledgement, and interrupting the conversation rather than contributing back to it.

Limitations of AI

In their desperate quest for AGI — an Artificial General Intelligence that can rival human intelligence — I fear that the frontier model vendors are making the same Jevons Paradox mistake as those enterprises that seek to apply AI merely to the knowledge they already handle. Built on the limited pool of recorded knowledge and conceived as an alternative to humans, these AGI-focused models ignore the untapped reservoir of talent, knowledge and real-world experience possessed by humanity at large. We advance as a species, not by knowing things, but by taking what we know and constantly reframing it.

In an earlier reflection on the limitations of AI, I noted that humans have learned through millions of years of evolution how to instinctively analyze the changing patterns around us and to select the optimal response for whatever situation we find ourselves in. That talent for adaptation has survived machines being able to beat us at chess and much more besides. I believe it will keep us ahead of attempts to reach the so-called singularity for a good few years yet.

Of course, these moments of knowledge acceleration in history — from the first written records, to the invention of printing, to the advent of the Web and mobile computing, now fueled by AI — are always hugely disruptive. When knowledge breaks out of its previous constraints, those who have prospered from their access to previously scarce knowledge lose out. They often fight a rearguard action to maintain that scarcity, rather than encouraging the new opportunities that open up. But history shows that it is those who seize the opportunities who come out on top.

What this means for all of us as individuals and as enterprises is that future success comes from embracing rather than resisting the surging velocity of knowledge. Futurist Jim Carroll has some good advice on knowledge velocity for individuals.

For enterprises, technology is taking the friction out of knowledge access and circulation. Those who prosper won't focus only on more efficient management of what they already know. They will seek out new reserves of knowledge, open up to new knowledge streams and networks, and ensure their people are present in those places where knowledge circulates most freely.