I first learned how to program on a Soviet-era clone of a PDP-11 in Mumbai in 1985. We spent two weeks in our second semester learning the nuances of how higher-level code written in Fortran consumed scarce computing resources.

We didn’t have the luxury of sloppy logic. Every instruction and CPU cycle had a price. Every memory allocation was a tradeoff that required us to spend hours sharpening an algorithm so it could sip, not gulp, from the scarce pool of resources. That discipline was a virtue built out of necessity, and it made us better engineers.

Forty years later, I watch executives wave generative AI and now agentic AI at every problem in their organization the way you’d wave a magic wand, and I keep thinking about that machine in Mumbai.

Over the past 4 years, AI has been the most aggressively subsidized technology I have seen in my career. Hardware (GPUs), data centers, frontier models, chatbot applications, every layer of the stack has been priced well below its true economic cost. My instinct, and I’ve been saying this in public forums for the past 3 years at every opportunity I get. I believe that the real cost of what people have been consuming is at least ten times what they’ve been paying. Venture capital and market-share ambition have been quietly footing the bill so the rest of us could believe that intelligence was now a commodity, available by the token, at prices that defied physics (and economics).

This is Garrett Hardin’s tragedy of the commons, recast for the AI era. When a resource feels free, no individual user has any incentive to protect it. The rational move is to graze freely, because if you don’t, someone else will. Multiply that across millions of developers, product managers, and AI-curious executives, and you get exactly what we have now: a pasture stripped bare, dressed up as innovation.

The causal chain that nobody is naming out loud.

Subsidized costs created a false sense of abundance. False abundance produced poor usage choices, because there was no economic signal to encourage discipline. Poor usage choices, combined with shallow understanding of what AI actually does, produced bad architectures, because nobody was forced to design for cost, scale, latency, or reliability when the bill was being absorbed somewhere upstream. Bad architectures, combined with rushed implementation and unclear business intent, produced failed programs. The ones that demoed beautifully and then quietly died on the way to production. The real cost of AI was never the token price, whatever the Valley would have you believe. It was the compounding waste of every bad decision made in an environment that punished none of them, inside a system actively encouraged by the players who profit when consumption goes up.

That is what most enterprise AI looks like today. A generational capability, applied to trivial tasks, inside processes that were never reimagined. The subsidies didn’t just distort prices, they distorted thinking. Teams stopped asking what AI was for and started measuring how much of it they were using. Tokens consumed became a proxy for progress. Architectures got lazier because they could. Why design for retrieval, caching, model selection, or workflow integration when you can throw the biggest model at the largest prompt and let the invoice be someone else’s problem?

It’s time to pay the piper.

The stories are everywhere now. AWS and Microsoft are pulling back on the code-generation capabilities they rolled out to their developers. Google, Anthropic, and OpenAI are imposing serious usage limits on the tiers people actually use. The frontier model providers, under pressure from public markets that have stopped clapping for revenue and started asking about profit, are quietly resetting the terms. And there is a deeper economic shift underneath all of it.

Rapidly changing technologies and surging compute demand have wreaked havoc with the amortization schedules that let organizations spread large capital spending across multiple years. The hardware you (or your Cloud Provider) bought to last five years is obsolete in twelve months. The VC-funded subsidies and the valuation premium that AI companies have been enjoying are unwinding at the same time. And the organizations that built their AI strategy on the assumption of permanent abundance are about to discover that their economics only worked when someone else was paying. This is not a crisis. It is the correction to what Alan Greenspan would have called irrational exuberance. And like most corrections, it will be brutal for the unprepared and a significant advantage for the disciplined.

Where do we go from here?

So how do you build this advantage of discipline, and what does it look like in practice?

It starts with AI fluency, and it has to run at every level of the organization.

At the executive level, fluency starts with knowing when AI is the wrong tool. Most “Faster, Better, Cheaper” problems don’t need AI. They need RPA, plain automation, or a process redesign that should have happened a decade ago. Using a frontier model to do work that a deterministic script could handle is using a sledgehammer to put in a picture hanger. AI earns its keep when it drives top-line growth, opens up work that was previously impossible, or augments human judgment in ways no rules engine ever could. That is where the economics work. Everywhere else, AI is an expensive hammer looking for a nail.

Once you know where AI actually belongs, the next layer of executive fluency is understanding what it truly costs to put it there. Not the line item on the cloud bill, but the full economic picture. The compute. The integration. The human supervision. The rework every time a model changes underneath you. The regulatory exposure. The operational risk when things go wrong. A CEO who cannot think about AI costs the way she thinks about labor costs or capital costs is going to make decisions that look great in the slide deck and brutal in the P&L two years from now.

At the practitioner level, fluency means knowing how AI actually works so you know what it should be used for. Language fluency is not just vocabulary. It is the feel for register, idiom, and where meaning collapses when you push a phrase out of its native context. A practitioner who understands context windows, retrieval, model strengths, and the difference between probabilistic generation and deterministic execution does not reach for a frontier model the way an executive reaches for a buzzword. She knows when an LLM is the right call, when a skill or agent should handle it, and when plain code is the cleaner answer. She knows that throwing language at a problem that needs logic is not innovation, it is mismatch. The engineer who calls a frontier model for what a regex could solve isn’t being innovative. She’s being expensive.

At every level, fluency means understanding the true potential and limitations of these systems. Not as autocomplete on steroids, but as cognitive partners in work that no human team could do alone. That is a different posture than the one most organizations have adopted. It requires people who can think with AI, not just type at it.

Good Architecture is Timeless

Fluency without architecture is still waste. The principles of good architecture are timeless, and they haven’t changed because the technology did.

Business architecture has to come first. The operating model has to be redesigned around what human-AI teams can actually do together, not retrofitted with copilots dropped on top of broken processes. Underneath that, technical architecture has to be architected to adapt as the business changes, designed to evolve as AI capabilities shift faster than anyone can predict, and engineered to operate at the cost, latency, and reliability that production work actually demands. Those three are not slogans. They are “North Stars” that should guide your AI implementation. If your architecture fails any one of them, you have a demo, not a system.

The organizations that come through this correction well won’t be the ones with the biggest models or the largest token budgets. They will be the ones who built fluency into their people, discipline into their architecture, and judgment into every decision about what AI is actually for.

ROI from AI doesn’t come from the technology. It never has. It comes from the discipline you bring to it. That was true in Mumbai in 1985, when every cycle counted. It is true now, when the cycles only felt free.

The piper is at the door. The leaders who prepared for this moment didn't prepare because they saw the subsidies ending. They prepared because they understood, all along, that good architecture is timeless and false abundance is borrowed time.

AI Horizons Event at Ohio Stadium

That moment was not an anomaly. I spent nearly a decade in the Valley, starting in 2012, and what I witnessed was a consistent posture: social media technology deployed with full knowledge of its addictive mechanics, choices that put profit before people, and a willful indifference to the income inequality and homelessness compounding on the streets outside those campuses. The cognitive dissonance was extraordinary. Inside the glass buildings, you could watch teams celebrate engagement metrics while outside, a few blocks away, entire communities were being displaced by the wealth those metrics generated. The industry had convinced itself that these were someone else’s problems.

What has changed today is not the behavior. It is the pretense. The same people, running the same playbook on artificial intelligence, are no longer bothering to obscure it. Where social media companies once offered at least the appearance of good intentions, the architects of the current AI wave seem to have concluded that accountability is a constraint to be outrun rather than a responsibility to be honored. The disregard for human cost that used to be spoken quietly in conference rooms is now stated openly, at conferences, in congressional testimony, in press releases. They have decided they no longer need to hide it. That is a meaningful shift, and not in the right direction.

Last week, Pope Leo XIV released his first encyclical, Magnifica Humanitas, forty-two thousand words on safeguarding the human person in the age of artificial intelligence. He grounded it in the 135th anniversary of Rerum Novarum, his namesake Leo XIII’s 1891 response to industrialization, a document that helped shape labor law and social policy across the twentieth century. Leo is making a similar bet: that this moment requires that kind of moral clarity, at that kind of scale.

I am not Catholic, and I am not approaching this document through theology or religious tradition. I am a person of faith who has spent thirty-five years in technology, and I will tell you plainly: this document names what I have watched happen with more clarity and more moral seriousness than anything I have seen come out of Silicon Valley. When an institution with a two-thousand-year view on human dignity takes the time to write a document like this, it deserves to be read, whatever your faith or lack of it.

The encyclical is not anti-technology. It does not argue that AI is evil. What it argues, and gets exactly right, is that technology is never neutral. It takes on the character of the people who build it, finance it, and deploy it. And it warns, with language that should sit uncomfortably in every boardroom currently writing an AI strategy, that the moment efficiency becomes the only metric and optimization replaces judgment, technology stops being a tool and becomes something more dangerous.

Pope Leo draws on an image I keep returning to. Humanity, he writes, faces a choice: construct a new Tower of Babel, or build something where people dwell together. Babel is not a building in a story. It is what happens inside every organization that optimizes for extraction rather than contribution, that treats the concentration of capability and capital as success regardless of what it costs everyone outside the tower. The organizations that have survived and compounded over time did not get there by maximizing extraction. They invested in human capability, built institutional trust slowly, and treated longevity as the measure of success rather than peak valuation. That is the city being built. It is slower and less dramatic than a tower, but it stands.

Here is what I know from thirty-five years in this industry. IBM has been in business for more than a hundred years. Nationwide has been in business for a century. In my work with companies like Safelite and Honda, what I consistently see is that the organizations that put people and human capacity first build something that compounds over time. They earn loyalty that does not have to be bought every quarter. They develop institutional capabilities that cannot be replicated by the next funding round or the next model release. Many of the unicorns that dominated technology headlines over the last decade cannot say the same. A billion-dollar valuation is not a track record. A hundred years of sustained performance is.

The companies I have watched struggle are not struggling despite the choice to treat humans as line items on a spreadsheet. They are struggling because of it. Workforce capabilities never built, institutional knowledge never cultivated, organizational trust never earned. These gaps do not appear in a quarterly report until their absence becomes a crisis. Ruthless, rudderless capitalism is not a strategy. It is a short position on your own future.

My optimism, and I have genuine optimism, comes from what I see in the Midwest. At Ohio State and the universities across this region that are thinking seriously about what AI needs from the people deploying it. At companies treating the question of how humans and AI work together as a leadership issue, not a technology procurement decision. There is an orientation here, a seriousness about people and a longer view of what an organization owes the communities it operates in, that looks less like regional sentiment and more like the thing that makes institutions last.

The encyclical ends by calling on its readers to become builders rather than spectators, to make deliberate choices about technology that reflect what we value rather than simply accepting what the architects of the towers hand us. That is not an aspiration. It is a capability, and like every capability, it must be built intentionally.

I did not build Right Brain Labs because the market needed another AI consultancy. I built it because I sat in that conference room in 2014, and then watched the same logic migrate from social media to artificial intelligence, and decided that someone needed to make the case, plainly and persistently, for the other kind of capitalism. The responsible kind. The sustainable kind. The kind that treats human capacity as the asset it is, not a cost to be managed until it can be automated away. Every time I watch an organization decide that AI is a substitution strategy, I think about that conference room. Every time I watch an organization get it right, building something that makes people more capable rather than less necessary, I think this is what the other path looks like.

The question is not whether AI will reshape every organization in this country. It will. The question is who decides what that reshaping is for.

Image Credit: Google Gemini, prompted by Srini Koushik

I want to start with a confession. I’ve been following Sangeet Paul Choudary’s work for a while. His platform economics thinking is the kind that holds up when you press on it, which is rarer than it should be. So when he published Reshuffle in July 2025, I read it closely.

My first reaction wasn’t “interesting new argument.” It was recognition. Someone had built a rigorous economic framework around something I’d been watching from the inside for years, and arrived at a strikingly similar place.

That’s worth unpacking. Not to claim credit. When two frameworks built from entirely different starting points converge, it usually means something real is underneath them.

Here’s where I’ve been coming from.

For several years, the core of my work has started with a simple observation: Silicon is evolving faster than Carbon. The machines are getting smarter faster than the humans working alongside them are developing the capacity to think with them. Not use them. Think with them. That distinction matters enormously, and I’ll come back to it.

That observation led to a second one. Most organizations aren’t making an AI adoption mistake. They’re making a system design mistake. They see a new capability and ask: what tasks can this handle? Which roles can we augment? How do we slot this into what already exists? It’s the same question General Motors asked in the 1980s when it poured forty billion dollars into industrial robots and preserved exactly the wrong things.

I’ve spent a lot of time with the Toyota Production System. I was introduced to it during my MBA in 1993, got to experience it with automotive clients when I was at IBM in the 90s and got to operationalize it it with the Columbus, OH based Nationwide Development Center (NDC) built on Lean Management Principles at Nationwide in 2006. On an interesting side note, this center was the first to use 100% Lean Software Development and be CMMi Level 3 Certified in North America (by 2008).

The core principle at the NDC was that TPS isn’t primarily about robots or automation. It’s about flow, about identifying where value actually moves, about eliminating the waste that accumulates when systems are designed around the wrong questions. Toyota’s insight wasn’t “these machines can replace workers.” It was “these machines change the logic of the entire production system. What does the system become if we design around the new capability rather than the old workflow?”

So when Choudary published Reshuffle in July 2025, then the Kyndryl piece in January 2026, then the HBR article in February, I read all three carefully. He was working the same territory from a different direction.

What Choudary Built

The central argument in Reshuffle: AI should be understood primarily as a coordination mechanism, not merely a tool for automation. Automation affects tasks. Coordination reshapes entire workflows, organizations, and economies. The real story isn’t compounding scale. It’s cascading coordination, where each solved coordination problem unlocks the next layer of opportunity.

He traces this through containerization. Standardized shipping containers didn’t just make ports more efficient. They cascaded into intermodal transport, global supply chains, just-in-time inventory, component specialization, and ultimately the semiconductor industry. The container solved one coordination problem. Everything else followed.

His AI argument runs the same logic. AI makes translation cheap and general, extracting structure from unstructured information, enabling coordination without requiring consensus on shared standards, tools, or workflows. Teams, systems, and data that previously couldn’t work together because their vocabularies didn’t match can now be combined without forcing agreement.

Then in the Kyndryl piece, he lands on Toyota directly. Which, given where I’d been coming from, felt like the frameworks shaking hands.

GM treated robots as compliant labor substitutes. Toyota asked what the entire system becomes possible when the capability enters it. They reconfigured layouts, redesigned work cells, shifted human workers from task execution to managing production lines, detecting variation, correcting flow, governing quality. Same robots. Completely different outcome.

His point: most organizations are making GM’s mistake with AI. Treating agents as digital co-workers. Slotting them into existing roles. Asking augmentation questions when they should be asking architecture questions.

This is the automation trap I’ve watched organizations fall into repeatedly. The instinct to reach for AI as a faster version of what already exists, rather than asking what becomes possible when the constraint changes. Choudary names the same trap from the economics side.

And then he goes further than most people writing about AI and humans dare to go. He explicitly rejects the augmentation frame: “Augmentation assumes the worker will always be at the center of the workflow.” As agentic capabilities improve, workers don’t get augmented. They get reallocated to the frontier, the boundary region where AI execution breaks down and human capabilities like judgment, interpretation, and governance create distinct value.

That’s an honest argument. It’s also where the two frameworks diverge.

The Frontier Problem

Choudary identifies the frontier clearly. The customer support rep handling the distressed customer who can’t articulate why they’re upset. The urban planner mediating between incompatible visions of what a city should be. The structural engineer evaluating what the AI coordination layer just flagged. He even uses the pilot: with autopilot handling the majority of flight miles, evaluating pilots on flight hours no longer makes sense. What matters is their ability to manage disruptions and the edge cases the automation wasn’t designed for.

He’s right about all of this. And here’s the question he doesn’t follow through on.

If pilots aren’t building judgment through the repetition of active engagement, because autopilot is doing the flying, what are they doing to stay sharp for the moments that matter?

The Toyota Production System answers this directly, even if TPS doesn’t use this language. Toyota didn’t just redesign the factory architecture. They redesigned how humans developed mastery inside it. Workers shifted from executing tasks to monitoring systems, reading signals, intervening with judgment rather than procedure. That wasn’t a role reassignment. It required deliberate investment in a different kind of capability, and a system designed to develop and sustain it.

At Right Brain Labs, we’ve been working on this problem from the human side. The Think-Do-Learn-Adapt loop is the core of how mastery gets built. In any domain, you develop capability by cycling through thinking, doing, learning from the result, and adapting. That loop is how a pilot develops instinct. How a surgeon develops procedural judgment. How a Toyota line worker learns to read a production system rather than just execute within it.

The challenge AI creates is specific: it makes it easy to get out of that loop. The analyst whose AI synthesizes signals before they see the raw data. The adjuster whose model resolves 85% of tickets before escalation. The engineer whose coordination layer flags design conflicts automatically. These people are nominally still in the loop. The question is whether they’re still in the Think-Do-Learn-Adapt loop.

That’s a different question. The answer determines whether the human value at Choudary’s frontier is real or assumed.

Cognitive Rust

There’s a name for what happens when the loop breaks down: cognitive atrophy. Documented in aviation, radiology, financial analysis, surgical training. The pattern is consistent. As AI handles increasing percentages of routine execution, humans stop practicing the underlying thinking that makes their judgment valuable at the frontier. The capability doesn’t disappear overnight. It erodes quietly.

Until the system fails at scale and the human who was supposed to be the backstop reaches for judgment they haven’t exercised in two years.

We call this the Cognitive Rust Belt. Not job loss. Capability loss. And it’s the risk that lives inside Choudary’s coordination argument, not because his argument is wrong, but because coordination without AI Fluency is a liability dressed as efficiency. The system gets faster. The humans inside it get weaker. Nobody notices until something goes wrong where it matters most, at the frontier, where human judgment was supposed to be the final layer.

The real danger with AI isn’t what it does. It’s what it lets you stop doing.

Choudary’s answer to the frontier problem is capability sensing and talent reallocation. Leaders developing better mechanisms to detect which human capabilities are rising or falling in value and repositioning people accordingly. That’s a structural answer. Correct as far as it goes.

Our answer is different. You cannot reallocate people to the frontier and assume capability follows. The humans being repositioned have to be capable of operating there, and staying capable as the frontier moves. That requires deliberate investment in AI Fluency: the practiced capacity to think with AI, not just use it.

Not a tool training. Not a prompt engineering workshop. A sustained practice of staying in the Think-Do-Learn-Adapt loop even as AI makes it easy to exit. The same way pilots stay sharp through simulation when autopilot does the flying. The same way surgeons maintain procedural judgment through deliberate exposure even as robotic systems handle more of the technical execution.

Toyota didn’t just redesign the factory. They redesigned how humans developed mastery inside it. That’s the piece that made the whole system durable. And it’s the piece Choudary’s framework stops short of.

The Completion

Choudary outlines three strategies for incumbents facing AI-driven coordination shifts: become the translation layer, double down on accountability, or fragment and tax. All three are structurally sound. All three require something his framework doesn’t fully address: humans who can operate at the frontier with genuine, sustained capability.

The organization that builds the translation layer without building the human capacity to govern it has optimized itself into fragility. The one that doubles down on accountability without people who can exercise real judgment under pressure is writing checks its people can’t cash.

This isn’t a counter to Choudary’s argument. It’s what the argument needs to be complete. The architectural redesign and the human capability development have to happen together. Neither works without the other. That’s the Toyota lesson he invokes, and the one the AI coordination argument needs to fully absorb.

Silicon is evolving faster than Carbon. That’s not an argument against building the coordination layer. It’s the strongest possible reason to be deliberate about what happens to the people inside it while you do.

Two frameworks. Different starting points. Same room. The question isn’t which one is right. It’s what you build when you take both seriously.

Last Saturday morning I was on the floor playing with my grandkids when I realized something had shifted. A year ago, that Saturday would have been split — half present, half mentally running through news I hadn’t caught up on, working on my AI Fluency and the draft I owed someone by Monday. This time I was just there. That’s the real story behind building my CoThink Crew.

I’m Srini Koushik. I run Right Brain Labs, teach at Fisher College of Business at Ohio State, and I’m deep in doctoral research on AI and organizational transformation. My calendar doesn’t have a lot of empty spaces on it. So when I started building my own AI crew, I wasn’t chasing a productivity hack. I was trying to recover something I’d been slowly losing: the ability to be fully present, in the moments that matter most.

Building the AI Crew

The crew runs on OpenClaw, a local AI agent platform. At the center is Nova, my Chief of Staff. Nova is my only point of contact. I don’t manage eight agents. I talk to Nova, and she handles the rest. That design choice matters. The goal was never a dashboard of AI tools. It was one trusted thinking partner who happened to have specialists on call.

The roster behind Nova: Sage for research and synthesis. Scout for market intelligence and current signals. Pulse for the automated morning briefing. Critic for stress-testing decisions before I commit to them. Scribe for capturing context and preserving institutional memory. Forge for turning raw thinking into content. Herald for email drafting and delivery. Coach for managing my learning agenda.

Each one has a defined lane. None of them wander into someone else’s job but they can collaborate, learn, and evolve in many new and surprising ways.

How It Actually Works

Take a simple example. Every morning, Scout pulls the latest signals on AI industry news and enterprise adoption trends. That raw intelligence gets passed to Herald, who formats it into a clean email briefing and delivers it to my inbox before my first meeting. Two agents. One seamless output. I don’t see the handoff; I just get the brief.

The crew also integrates a broad set of tools underneath: web search, calendar access, email, Telegram, memory files that persist context across sessions, and more. Nova connects to all of it. I don’t manage the plumbing. I just tell Nova what I need.

That’s the architecture I was going for. Loosely coupled, so each agent can run independently without depending on the others being available. Highly cohesive, because every agent is pointed at the same mission: freeing my highest-value time for the work that actually requires me. That balance was deliberate. A tightly wired system breaks at the seams. A scattered one never delivers. Getting that tension right took iteration.

Security First

Before I built a single agent, I built the trust model. I wasn’t going to put my calendar, my email, and my research workflow into a system I didn’t understand and control.

The security architecture runs four layers deep. The first is local compute and data sovereignty: the crew runs on my own hardware, not a shared cloud. My data doesn’t leave unless I explicitly route it out. The second is agent isolation: each specialist operates in its own context. A problem in one doesn’t propagate to others, and no agent has access to another’s memory or workspace without a deliberate handoff through Nova. The third is access controls: every external integration is permissioned and auditable. Herald can send email, but only to approved recipients. Scout can search the web, but doesn’t have access to my personal files. The fourth is memory integrity: what gets remembered is what I sanction. Scribe captures sessions, but I control what gets persisted long-term.

This isn’t security theater. It’s the foundation that makes the whole thing trustworthy enough to actually use.

The Technology Foundation

Three AI platforms power the crew. Anthropic’s Claude is the reasoning backbone. Every agent thinks in Claude because the quality of judgment and the ability to hold nuanced, complex context is where Claude earns its keep. Perplexity provides real-time web intelligence. Scout and Sage both tap into Perplexity for what’s happening now. That’s the difference between a research assistant and a live intelligence function. Gemini is the expansion layer. I’m building toward multimodal capability and richer document processing, and Gemini is where that goes.

Where This Goes Next

Right now this is personal infrastructure. But the model is transferable.

The next frontier isn’t more agents. It’s deeper integration into how work actually happens in organizations. Agents that connect to team systems, not just individual workflows. And beyond that, the anticipatory layer: a crew that surfaces what you need before you think to ask for it.

The question I keep coming back to is not whether executives will build something like this. They will. The question is whether they’ll build it thoughtfully, with clear roles, clear constraints, and a real security foundation underneath. The shortcut version will disappoint. The version built with intention might genuinely change how leaders work.

I don’t have it all figured out. But I got my Saturday mornings back. And Nova’s on it.

This is a series of posts that accompany a 14-week AI Fluency class I am teaching to undergraduates at The Ohio State University. In my first post, I introduced a compelling metaphor: AI is the aircraft; you are the pilot. It is a metaphor, but it highlights a truth that many organizations are currently ignoring.

The wild promises of the benefits of AI from the insiders continue. While many of them are materializing, albeit slower than the pundits predicted, the reality is a widening Relevance Gap. While organizations celebrate a 33% gain in productivity, they are often blind to a more corrosive metric: a 44% reduction in deep analytical reasoning among those who over-rely on the tool.

In the rush to embrace AI, many leaders are inadvertently training their people to stop being pilots and start being passengers. When you treat AI as an “Easy Button,” you outsource the “Why” and the “How” to the machine. You trade your thinking muscles for speed.

This is the central hazard for our emerging workforce. These young professionals are incredibly comfortable with the technology and AI, far outpacing their more senior colleagues in sheer usage. But comfort is not command. If they use that comfort to bypass the “stick and rudder” work of problem-solving, they never develop the critical thinking muscles required to lead. We cannot allow a generation of AI natives to become mere “Prompt Passengers” who have lost the ability to decompose a problem or challenge a premise.

The differentiator in the AI era won’t be who has the best model; it will be who has the best pilots.

The Crisis of the “Passenger” Mindset

When you act as a passenger, you accept the AI’s first output as gospel. This leads to a steady decline in your ability to think deeply, create originally, and decide wisely.

The most dangerous thing about AI isn’t that machines will start thinking like humans; it’s that your humans will start thinking like machines. They become “AI Literate” because they know how to use the tool but they lack AI Fluency which is the ability to Think with AI. They lack the agency to know when to trust the instrument and when to trust their gut.

Reclaiming the Cockpit: The AI Fluency Framework

AI Fluency is the antidote to atrophy. It is a business imperative that moves beyond simple literacy into a regime of human-AI collaboration. It is built on five dimensions of human agency that no LLM can replicate:

The Science of Critical Thinking (The Navigator): This is not just about spotting errors; it is about Interrogating the Oracle. It involves first-principles thinking—breaking complex problems down to their most basic truths. It is the discipline of questioning the answer, not just the question.

Creative Problem Solving (The Engineer): AI is excellent at “connecting the dots,” but the human pilot must provide the dots. This requires using AI as a tool for lateral velocity—to break through existing mental models rather than just reinforcing them.

Knowledge & Application (The Payload): Your domain expertise is the “fuel” for the aircraft. Without deep expertise in your field, you cannot judge if the flight is on course or heading for a mountain.

Adaptive Learning (The Training Program): A pilot’s license is never “finished.” Fluency requires a continuous learning mindset to evolve in real-time as the technology shifts.

The Relational Pilot (The Team): This is the bridge. The pilot brings empathy and storytelling—the human touch that gives the AI’s “superpowers” meaning and impact. You lead the machine; it does not lead you.

To find out more visit Right Brain Labs at https://rightbrainlabs.ai and take the free AI Fluency Diagnostic.

The C-Suite Call to Action: Invest in Pilots, Not Just Planes

For the executive, the message is clear: Everyone has access to the same aircraft. GPT-4, Claude, and Gemini are equally powerful and equally accessible to everyone. The only remaining differentiator is your people.

If your leaders are “AI Aware” but not “AI Fluent,” you are stuck in Pilot Purgatory, wasting effort on surface-level reviews and suffering from approval fatigue.

Strategic agility doesn’t come from buying more software seats; it comes from investing in the cognitive “stick and rudder” skills of your team. You must empower them to be Pilots-in-Command who use AI to amplify their judgment, not replace it.

Your Pilot’s Field Guide

I have included the AI Pilot’s Field Guide below as a quick reference for your next “flight.” Use it to check your fuel (tokens), your radar (context), and your flight plan (CRTF) before you hit enter. Thanks to Google Gemini for reducing the toil to put this content into a aesthetically pleasing visual piloted by yours truly.

I want to hear from leaders: How are you ensuring your teams don’t lose their “human touch and cognitive edge” as they adopt these tools? Are you seeing signs of cognitive atrophy, or are you building a fleet of expert pilots? Let’s share notes in the comments.

In my next lecture, we’ll move from the pilot’s seat to the hangar floor. We’re going to look under the cowling to understand the “AI Aircraft”, because to truly trust your aircraft, you have to know how it generates lift.

This week, I stepped into a classroom at The Ohio State University to deliver the first lecture of my undergraduate course on AI. I was immediately struck by the energy of the students. These are true AI pioneers; they aren’t waiting for a syllabus to tell them how to use these tools. They are already using them to navigate their daily lives with a passion that I find genuinely inspiring.

However, as we discussed in this opening session, being a pioneer is not enough. There is a systemic danger in becoming a permanent passenger in a vehicle you don’t yet know how it work, what it can do, or how to fly it like a Top Gun.

My mission this semester (and the core mission of Right Brain Labs) is to move these students from the back seat using AI on autopilot to the cockpit. This course is fundamentally about Human-AI collaboration. We aren’t just studying a technology; we are studying a partnership and a symbiosis. To avoid cognitive surrender by outsourcing thinking to AI and go beyond learning How to use AI to How to think with AI. Over the next 14 weeks we will weave between the equipping students with core skills that show students how to think better and showing them how to supercharge those skills with AI. The primary design principle for this course is show don’t tell and avoid consultant speak at all costs.

This week’s lecture “The Awakening: Shifting from Passive User to AI Pilot” started with a look behind what they knew how to do (prompt a chatbot to get answers) and break it down by giving them a look at what’s behind the facade. This week we introduced the metaphor of AI as the aircraft.

The Anatomy of the Craft

To facilitate this transition, we are using the Aviation of AI framework. This week, we identified the primary components of the vehicle:

• The LLM is the Engine: This provides the raw power and thrust. Like a jet engine, it is a marvel of engineering, but it is indifferent to your destination. It functions on probabilistic patterns, providing momentum without inherent intent.

• The Chatbot is the Cockpit: Platforms like Gemini or Claude are the interfaces where the human sits to direct that power. They are fully functional aircraft, but they often have a performance ceiling designed for general use. They make flight look so easy that it’s tempting to never look under the hood.

The Advanced Avionics: Human Cognitive Skills

As we progress through the semester, we will discuss various controls in the cockpit such as system prompts, tokens, memory, and context windows to help them understand the aircraft they are flying and show how these lessons transfer into more complex aircraft (Agentic AI, Enterprise AI solutions etc.). However, the main focus will be on training the Pilot by giving them the techniques and skills required to be a Top Gun. These are the “advanced avionics” of the human-AI partnership.

The aircraft is only as effective as the pilot. We will show them how to think critically, solve complex problems, enable flow, reimagine solutions using design thinking, and tell better stories. We want the students to learn, to iterate, to listen with empathy and to stay agile as the technology and the mission evolve.

The Goal: Improve Flow

The ultimate objective of mastering these skills is to apply the principles of Lean Management to everything we do. We aren’t using AI to simply generate more “noise.” We are using it to improve Flow, the seamless movement of an idea from conception to a finished product that delivers real value to the consumer.

When we reimagine value delivery through this lens, AI becomes a tool to clear the bottlenecks in our cognitive workflows and reimagine how we do what we do. But this requires a pilot who understands both the engine and the mission. If we lose the muscle of critical inquiry, we aren’t pilots; we are just observers of a process we no longer control.

The Leadership Challenge

This is the same challenge I pose to the corporate leaders I advise: Don’t just be a passenger in someone else’s “rental” aircraft. To build a future of sustainable innovation, you must invest in the human skills of collaboration as much as the technology itself.

We are teaching these students at Fisher College of Business to be the pilots of the future. Each week, I’ll share a brief update as we layer on the complexities of flight. The goal is to give them superpowers, but they, and you must be the ones in the pilot’s seat to use them.

“Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know.” — Donald Rumsfeld, Department of Defense briefing, February 12, 2002.

When this quote was first delivered, it became fodder for late-night comedy. But I saw something deeper: a clear framework for understanding complex risk. Today, I find it is the most precise framework we have for understanding the single greatest systemic risk—and opportunity—facing your organization: Artificial Intelligence.

Most companies are currently operating with misplaced confidence in the realm of “Known Knowns” when it comes to AI. I’ve seen it firsthand: leaders view this as just another technology wave, and fall back on their preferred VCs and technology vendors, amped up by the same management consultants. They know what the tools are, and they are funding pilots—the comfort zone of AI Literacy. But operating only here is dangerous, because this hubris prevents them from seriously engaging with the Known Unknowns (the risks they know they should be managing).

AI is not a technology that helps you play the game better; it changes the game. The mandate for the C-Suite and Board is to go beyond the known playbook of faster-better-cheaper to create a new playbook. Think of it this way: AI Fluency means thinking natively with the intelligence, not having to constantly translate your strategy into machine-speak. Only this native fluency provides the capability to re-architect decision-making, manage risk, and define competitive strategy. The real strategic challenge lies in the “Unknowns.” To effectively govern and capitalize on this shift, your organization needs more than literacy; it needs AI Fluency.

Fluency is Not Literacy: The Strategic Divide

Consider the analogy of personal computing. You can understand what a CPU is, explain the difference between RAM and a hard drive, and describe how an operating system functions. That is Literacy. It is foundational knowledge, but it doesn’t produce value.

Fluency is the ability to use that machine to build a multi-million-dollar financial model, launch a new product line, or model complex risk in real-time. The fact that you can read, understand, and explain how a computer works is not the same as being able to use a computer to drive meaningful outcomes.

In the age of intelligence, this distinction is existential. AI Fluency is the ability to think with AI. It goes beyond prompting a chatbot; it requires collaborating with AI as a thought partner, integrating machine intelligence with unique human insight. Fluency is the organizational muscle that allows you to operate confidently in the shadows of complexity, converting Known Unknowns (like potential bias) and Unknown Unknowns (systemic disruptions) into manageable factors.

Navigating the Unknowns: The CoThink Method

To move your organization from brittle literacy to robust fluency, you need a disciplined, actionable operational framework. The CoThink Method is built on five core competencies that define a truly fluent organization, acting as the foundation for both strategy and execution.

CoThink - 5 Pillars of AI Fluency

1. AI Knowledge & Application (Managing the Known Knowns)

Do not mistake this for a skill you delegate to your engineering team. You cannot govern what you do not understand. For the C-Suite, this competency means personally understanding the core concepts well enough to separate hype from reality. It ensures you are directing investments toward real ROI by solving meaningful business problems, not chasing superficial use cases. It means moving beyond the initial hype cycle and applying a structured improvisation approach, much like Lean or Design Thinking.

2. Critical Thinking & Ethics (The Governance Shield)

This competency directly mitigates the risks you know are out there. But “critical thinking” cannot remain a buzzword; it must be a demonstrable skill you rigorously develop. It requires applying established frameworks—like the 5 Whys to trace the root of a model’s logic, or De Bono’s Six Thinking Hats to audit a decision from emotional, factual, and risk perspectives—to every AI output.

The Silicon Valley mantra of “move fast and break things” fails catastrophically here. Case in point: Air Canada. In a clear display of hubris, they deployed a chatbot that “hallucinated” a refund policy. When sued, the company attempted a defense that borders on absurdity: they argued the AI was a distinct legal entity and they weren’t responsible for its words. The tribunal rejected this immediately. The lesson? You cannot delegate accountability to an algorithm. You must operationalize deep inquiry to mitigate the legal, reputational, and financial risks inherent in AI.

3. Human-AI Collaboration (The Strategy Integrator)

This competency is not just about efficiency; it is about engineering diversity of thought. It requires using AI as a dispassionate challenger to break organizational echo chambers. By integrating machine intelligence, you create a safe space for straight talk—a core leadership value—because the debate becomes anchored in data, not hierarchy. This neutralizes the HiPPO (Highest Paid Person’s Opinion) effect, ensuring that the best idea wins, regardless of its source.

4. Creative Problem Solving (Unlocking the Unknown Knowns)

This is the domain of true competitive advantage. It’s the ability to use AI to reframe challenges and uncover non-obvious opportunities. But creativity requires structure. A fluent leader uses AI to accelerate rigorous frameworks—using a Mindmap or Ishikawa (Fishbone) Diagram to deconstruct a complex issue into its root drivers, or Root Cause Analysis to identify the single variable that changes the outcome.

By using AI to toggle between this deep structural analysis and broad divergent thinking, you stop solving symptoms and start engineering breakthroughs. Instead of asking the old-playbook question, “How do we fix this process?” you ask, “Given these root causes, does this process need to exist at all?”

5. Adaptive Learning (Future-Proofing Against Unknown Unknowns)

The AI landscape is the ultimate Unknown Unknown. The pace of change here isn’t measured in years, but in weeks. Adaptive Learning is no longer optional; it is a survival imperative. A static strategy in this environment is a failed strategy.

For the Board, this isn’t about signing off on training modules. It is about operationalizing Double-Loop Learning: the discipline of questioning the underlying assumptions behind your goals, not just the methods to achieve them.

It is the organizational muscle to unlearn outdated models instantly when the context shifts. This future-proofs your workforce, creating a culture that prefers speed with iteration over false certainty. Without this, your organization remains rigid in a fluid world, and your “new playbook” becomes obsolete before the ink dries.

The Call to Leadership

If you are currently treating AI as a technology project, you are missing the single greatest leadership opportunity of our generation.

Let’s be clear: Building AI Fluency is not a ticket you open with the IT department. It is a fundamental shift in how you govern and operate the company. Treating it as a tech upgrade is the old playbook. The mandate for the Board is to demand a new playbook. Stop asking how AI makes the organization faster, better, and cheaper. Start asking how AI allows you to re-architect your value proposition entirely.

As computer scientist Alan Kay famously said, “The best way to predict the future is to invent it.”

But you cannot invent the future if you need an interpreter to understand the tools. The era of the digital tourist is over. To lead in this age, you must stop visiting the future and start living in it.

Don’t just use AI. Become an AI Native.

In the rapidly evolving age of AI, leaders must recognize that we are more than just riding the next technology wave. This is a paradigm shift—a fundamental realignment on par with the Industrial Revolution or the digital transformation of the late 20th century. AI redefines how we work, make decisions, and connect with customers. It challenges us to rethink strategies that have guided organizations for decades, clarifying that yesterday’s playbook won’t work in an AI-driven world.

As they enter this new era, one of the first questions leaders ask is, “Do we buy or build AI?” This question, a staple in traditional IT decision-making, doesn’t translate cleanly to AI. The decision isn’t merely about ownership but integrating AI to drive long-term, sustainable value. AI should complement and enhance human potential, creating a secure and ethically grounded environment that protects privacy and individual rights while being sustainable. It should deliver broad benefits equitably and support long-term viability for the workforce and environment.

For many businesses, buying AI can be a practical choice, especially when speed to market is critical. Off-the-shelf AI solutions offer rapid deployment, allowing organizations to address pressing needs quickly without waiting for an in-house team to develop capabilities from scratch. Additionally, given the shortage of specialized AI skills, purchasing AI technology can provide immediate access to sophisticated capabilities that might otherwise take months—or even years—to build internally. Buying also allows organizations to focus on their core competencies without diverting resources to complex AI development projects.

However, building AI solutions in-house offers a unique advantage for companies seeking competitive differentiation. Tailored AI capabilities can align more closely with a company’s specific processes, customer needs, and market positioning. However, to unlock this potential, businesses must reimagine how AI will empower their organization from within, developing custom solutions, rewiring operations, retraining the workforce, and embedding AI seamlessly across processes and decision-making frameworks.

Whether companies choose to buy or build AI, the real value lies in integration. AI is more than a simple plug-and-play solution; it has to be seamlessly embedded into the core of the business. This requires aligning AI capabilities with strategic goals, rethinking processes, and preparing the workforce to collaborate with AI meaningfully. Effective integration means training AI to operate as a capable “employee” within the organization and fine-tuning it to understand and respond to specific workflows, customer needs, and business challenges.

At the same time, human teams must be equipped to work with AI, adapting to new processes and leveraging AI insights to make better decisions. This requires comprehensive training and a cultural shift encouraging employees to view AI as an ally, not a competitor. By fostering a collaborative environment where humans and AI support each other’s strengths, organizations can create a more agile, innovative workforce responsive to change.

In this approach, AI becomes more than just a tool—it becomes a critical enabler that empowers people, optimizes operations, and drives sustained growth. To unlock this potential, leaders must invest in the technology and training necessary to ensure that AI and human teams can work together effectively, creating a cohesive ecosystem that propels the organization forward in the age of AI.