Dan Priestley shared a prediction yesterday that AI will trigger a financial meltdown by 2029. The argument is that we're spending hundreds of billions on infrastructure—data centers with a three-to-four-year lifespan—with no financial model to justify it. I am not a financial expert (I switched from microeconomics to business informatics for a reason), but the argument sparked something.
Because, for me, buried inside that argument is something important that is far less discussed: Every major infrastructure wave in the last 180 years has not just reshaped the economy. It rewrote the rules of knowledge — who owns it, who can access it, and what happens to those who don't adapt to it.
AI will do the same. But this time, the rewrite is personal.
Before railways, knowledge was local and embodied. The blacksmith knew his craft. The farmer knew his land. Knowledge lived in people and communities, passed through apprenticeship and conversation. Railways created standardized time (before railways, every town had its own clock), mass print culture (newspapers could distribute nationally overnight), and the professional class — lawyers, accountants, and engineers who sold codified expertise across geographies.
Communities whose knowledge was purely local got economically bypassed. If you only knew how to serve your town, and the railway now meant your town competed with every other town, your local knowledge was suddenly worth less. The winners were the people who could transport their expertise: write the book, give the lecture tour, and consult across regions. Local mastery wasn't enough. You had to be able to move it.
Electrification didn't just power factories. It enabled the modern office — the typewriter, the filing cabinet, the telephone. With them came the first documented, organized corporate knowledge: Standard Operating Procedures, org charts, and job titles. Frederick Taylor's "Scientific Management" movement literally extracted knowledge from workers' heads and encoded it into management systems. The goal was efficiency. The consequence was that individual craft knowledge became industrialized, and workers who held that knowledge became interchangeable.
If you were an executor (someone who did the work), you were at risk of being replaced by someone who could follow the system you used to carry in your head. The winners were the people who could design systems, not just operate within them.
The highway network didn't just move people faster. It made the franchise possible. McDonald's, Holiday Inn, Walmart. These weren't just businesses. They were knowledge compression technologies. The proven way to run a food operation, a hotel, a store — standardized, documented, and replicable across geography. Alongside this, television created a one-to-many knowledge broadcast: a tiny number of voices decided what millions of people knew.
Deep, contextual, local expertise was devalued again. The local diner didn't lose to McDonald's because the food was worse. It lost because McDonald's had turned the knowledge of running a restaurant into a system anyone could follow. The winners understood replication and model design. The people who thrived weren't just good at their craft — they knew how to package it so others could execute it.
The internet did two contradictory things simultaneously. It democratised access to information — anyone could know almost anything. And it created catastrophic noise — most of what was findable was low-quality, misleading, or outright wrong. The people who got wrecked were those whose value was in knowing things: travel agents, encyclopedia salespeople, researchers, print journalists. The internet commoditized their stock-in-trade overnight.
If your identity was "I have access to information others don't," that identity evaporated. The winners understood that in a world of infinite information, the scarce resource is no longer information itself. It's trusted judgment — who to listen to, what framework to apply, whose lens to trust. This is when "personal brand" stopped being a vanity concept and became a survival concept.
Each previous wave commoditized something that was once scarce, thereby creating new scarcity.
AI is commoditizing trusted judgment itself. This is the part most people are not taking seriously enough. For the last fifteen years, the winning move was to build a voice, a platform, a personal brand. Publish your thinking. Establish your authority. Become someone worth listening to. That advice was correct — for that era.
But AI can now generate plausible, well-articulated expert content at infinite scale, on virtually any topic, in virtually any voice. Which means the content layer — the blog posts, the frameworks, the LinkedIn articles, the podcast takes — is rapidly becoming as commoditized as information itself was in 2005.
The inversion is coming. The more AI generates, the less any individual piece of content can be trusted. The more voices that exist, the harder it becomes to know which ones are real. And at a certain threshold — probably 2026 to 2028 — trust in any content will collapse unless it can be verified as genuinely human, consistently expressed, and tested over time.
Here is the pattern that holds across every single wave: The more a wave produces of something, the more valuable its authentic opposite becomes.
This isn't a business observation. It's almost an unwritten rule. And it means the window to establish yourself as genuinely, verifiably you — before the collapse makes it nearly impossible to be believed — is right now.
There is a second dimension to this that matters just as much. This is an uncomfortable problem. It's kind of like eating your vegetables. Everyone knows that to be healthy, you need to exercise and eat well. Few do it consistently. Similarly, everyone in the knowledge economy knows they should document their expertise, articulate their frameworks, and build an accessible body of work. Almost nobody does it — because it is genuinely uncomfortable. Being truly known means having your real thinking on record, not the polished version. It means being legible, and therefore vulnerable.
At some point soon, AI will solve the friction in knowledge extraction. Ambient tools will begin passively observing how an expert thinks, decides, and communicates, and will construct a knowledge model without requiring the expert to sit down and articulate it. The technical problem of extracting knowledge will be largely automated by 2027. But automation can extract. It cannot confer legitimacy. It cannot verify authenticity. And it absolutely cannot do the deeper work of helping someone discover who they actually are at their most essential — and commit to living from that place. That work remains irreducibly human.
If you run a B2B company — say, somewhere between $5 and $30 million in revenue, a complex solution, buyers who take months to decide — you are probably already feeling the early edges of this shift if you are using AI. Buyers are arriving at conversations better informed but harder to impress. Your blog articles and whitepapers that used to feel distinctive are starting to sound like everyone else's. Your team is experimenting with AI tools, getting inconsistent results, and no one is quite sure whose job it is to fix them. What history tells us is that this feeling is not a tool problem. It is a knowledge problem. And it has a specific shape.
The knowledge that makes your company genuinely good at what it does almost lives in scattered places right now. In the founder's head. In the instincts of your best salespeople. In proposals rebuilt from scratch each time. When AI tries to represent your company, it has nothing real to draw from, so it generates something plausible and generic. Generic, in a long sales cycle with a committee of buyers, is a credibility problem you cannot afford.
"If you haven't documented your specific way of thinking — your frameworks, your worldview, your proprietary process — AI will replace the real you with a generic version of you by filling in the blanks (as it does for everyone else)."
The first stage of work is to change that: extract what good actually looks like in your business, codify it into standards your team can use and your AI can accurately draw from. This is not glamorous work. But without it, every AI tool you adopt will quietly erode buyer trust at a speed you cannot manually catch.
The companies that do this in the next 12 to 18 months will compound that advantage. The ones that don't will spend more to produce less convincing output. That said, this stage is a foundation, not a destination. As AI tools improve, the basic work of organizing knowledge will become cheap for everyone. The competitive advantage it creates today will narrow. Which leads to the harder question.
Once your knowledge is scaled and accessible, the real question is: does anyone believe it? In a market where every competitor is producing well-structured, consistently expressed, AI-assisted content, buyers will stop asking "do you seem credible?" and start asking "can I verify that you actually are?"
Trust at this stage is not a question of positioning. It is an evidence question — built through client outcomes, buyers can interrogate, through consistent positions held over years, through communities of people who vouch for you without being prompted. The practical shift for a B2B company is that marketing and sales converge around a single goal: making trust visible before the conversation begins. Not convincing buyers in the meeting. Educating them so thoroughly before it that they arrive having already decided.
This is the stage most companies are not yet thinking about — but it is worth naming now, because the inversion principle points directly to it. When AI has commoditized the content layer entirely — when every competitor has documented knowledge, consistent output, and visible proof — the final frontier becomes something that cannot be systematized. It is the specific way your company sees the world.
The underlying beliefs about what your clients deserve, what your industry gets wrong, and what doing this work with integrity actually requires. That worldview, when it is genuine and consistently lived, is something no AI can approximate and no competitor can replicate without becoming you. The companies that arrive here are no longer competing in a category. They have become one.
The inversion principle holds across every wave: the more infrastructure produces of something, the more valuable its authentic opposite becomes. AI is producing knowledge, judgment, and expertise at a scale that would have been unimaginable five years ago. Which means the scarce resource — the thing that will command attention and premium in 2029 — is verified human truth. The specific, tested, consistently expressed reality of who you actually are and what you actually stand for.
Every wave has rewarded the leader who saw that inversion coming and moved toward the new scarcity before it became obvious. The window to move is open. The question is whether you act on it before the companies beside you do.