I recently worked with a leader who, in every conceivable way, is tack-sharp, smart, and utterly capable. She’s the kind of leader that every organization wishes they had. A role model that her organization genuinely looks up to. Because of her star status, this leader had found herself under enormous pressure from executive leadership to show that she was leading the way on AI adoption in her organization.

In a show of that leadership, she had an important presentation coming up, and she decided to use AI to build a slide deck. She gave it a rough outline and loved what AI came back with. On the surface, it was astonishingly polished and thorough. “I can’t believe the number of hours AI has saved me,” she proudly shared with me. “Building slide decks is the bane of my existence.”

A week or so passed, and the time to deliver her presentation came and went. We spoke the day after. The expression on her face spoke volumes — utterly crestfallen.

She recounted how she walked to the podium, confident that she had everything well-in-hand. And how it all started to come apart from there. She wasn’t sure how to transition from one slide to the next. Each slide was so full of messaging that she wasn’t altogether clear what she wanted to say in the moment. And her delivery fell well beneath what she knew she was capable of.

She asked me what she could have done better, whether there was something she should have prompted the AI to do that she missed.

What I shared was that her workflow deserved a full look. When asked what she had done following the preparation of the slide deck, we got into what actually got missed: the refinement to make the presentation hers.

See, in having AI build the deck, the friction of deciding what deserves to be on each slide, what needed to be cut, and what she really wanted her audience to take away, was never addressed. She had skipped the most important part of public speaking preparation: rehearsal.

AI had automated the preparation, and in so doing, it also let her skip the work.

What got automated wasn’t the task

I share this story because this leader’s experience with AI has become increasingly common, and it’s become one of the reasons why some leaders have come to believe that “AI slop” is the only kind of output that AI is capable of. Not necessarily because the outputs are bad, but because the outputs were deployed without the human work that should have surrounded them.

Many organizations are eager to adopt AI in the belief that “saving time” or “eliminating manual work” through automation is the key to optimizing productivity.

AI is indeed very good at removing friction. What it can’t tell you is whether that friction was doing something you needed.

Business leaders often carry the intuitive view that friction is where the waste is. Eliminate the friction and we eliminate what keeps the business from marching forward. And sometimes that’s true.

What might be counterintuitive is that there are times when friction is actually load-bearing.

Friction is also where understanding gets built, and how sound judgment is formed. Where leaders stay connected to work in ways that matter. When that friction is automated away, you might save time. But, you might also lose something in the process — and you often won’t know what’s gone missing until you need it.

At Brilliant, I still review business expenses by hand every month. Not because I couldn’t automate the analysis; I could, and I do, with AI. But I do set aside the time to look at every line myself and with my team, and we ask: what was this for? Did it make sense? Should we keep doing it?

Some might say that process is inefficient. But, it’s also how I stay connected to the judgment calls we’re making as a business. An automated summary might give me the numbers, but it’s the manual review that gives me the understanding.

These are different things.

It’s “AI-first,” not “AI-everywhere.”

“Compliance” is how I’d describe what many organizations I encounter are doing with AI right now.

And I don’t mean in the regulatory sense of the word. What I mean is that organizations are responding to their Boards, who are asking about AI, and their senior leaders who are mandating it.

Somewhere in the middle are a lot of capable people who are nodding, executing, and performing visible alignment to AI, all while privately wondering whether any of it is producing what they were told it should.

Spend enough time with me and you’ll hear me remark on a distinction that I first heard years ago and come back to often:

Leaders do the right thing. Managers do things the right way.

In the context of AI adoption, a lot of people who ought to be leading are managing instead. They’re executing the directive well, but they’re not asking whether the directive is right.

That’s a distinction with an organizational cost because it’s one that goes beyond culture and influences final results.

But isn’t AI different?

I know what some readers will be thinking at this point.

But AI is transformative. It’s everywhere. I’m using it all the time. You can’t afford to hold back.

And in a way, you’d be right. Consider the Internet. In 1995, it was hard to build a financial model that justified Internet adoption. The ROI wasn’t always obvious. The metrics didn’t exist. And yet the organizations that held back because they couldn’t see the numbers paid for it for decades.

But unfettered ubiquity wasn’t how organizations adopted the Internet.

Those companies who found the most success in the dot com era weren’t necessarily the ones who put the Internet everywhere, into every process, as fast as possible. They were the ones asking a much harder question: where does this genuinely change what we can do, and where doesn’t it?

That’s a question that demands deliberation. It requires leaders to say “Here, and not here. This, and not for that.” Even when the boardroom wants to know how much and how soon.

AI is no different. One is leadership and the other is management.

What does AI leadership look like?

Many accountable AI leaders will be asked in the days ahead, if they haven’t already, to show how AI is driving business results.

And that’s where some important AI decisions will need to be made. It could be tempting to think that the decision is where AI must be most urgently deployed. But, I’d suggest that the most important question is actually the converse: where must AI deliberately not?

Don’t look at this stance as a retreat from AI, because it’s emphatically not. It’s actually the exercise of strategic judgment in its most powerful form. Not what an organization wants to do, but what it won’t do.

It starts with leaders being honest, with their organizations and with themselves, about the difference between what AI can really do to drive results and where it just gets in the way of achieving results.

Call those places where AI gets in the way your organization’s “no-fly zones.” These are the processes, practices, and decisions where deliberate human judgment produces something that speed and automation can’t replicate. And not because AI can’t replicate the outputs, because it often can. But because the act of doing it yourself is doing something that matters beyond the output being created.

Good leaders have clear no-fly zones for AI. They can name them, defend them, and they’ll resist the pressure to automate them away in the name of moving fast. They have the courage that every organization desires, even if they don’t know it yet.

You already know which parts of your work matter most when you do them yourself. You already sense where your organization is going through the motions. You already have a view on what would be lost if certain things got handed to a model.

That instinct is worth something, maybe more than you know. Act on it.

AI, Upfront publishes every Monday. If this was useful, subscribe to get it in your inbox. And if there’s a topic you’d like me to tackle, send me a message and let me know.

Subscribe now

When I meet with senior leaders to talk about AI for the first time, among the things I ask is what AI models their organization uses. Most will name one without hesitation. When I ask them why, the answers invariably are: “It’s the one I’ve gotten used to” and “it’s the one that gives me the best answers.”

If I ask them what they’re looking for in an AI model, they’ll often tell me they want “the best” model, or the “most accurate” model, or the one that makes “the fewest mistakes.”

Some tell me they keep a close eye on LLM leaderboards, in the same breath sharing they’re constantly concerned about doubling down on a particular model because it seems like there’s always a new one right around the corner, and the next model always promises a meteoric improvement over the last.

At the same time, I’ve run into leaders who mourn the loss of models they’ve used for a long time. I’ve spoken with leaders who were genuinely upset when an AI provider deprecated a model version they’d grown to rely on. As if they’d lost a trusted colleague.

One described to me that it felt like their assistant had been replaced overnight without warning. And the new one “just isn’t as good.”

These reflections reveal how we relate to the AI tools we invest in. They also reinforce how many of the decisions we make about AI are based on subjective experience, gut feel, and what we hear from others.

Changing AI benchmarks

Despite the hiring rigor that many organizations adopt, it’s curious that many leaders today don’t set out clear qualification criteria for what they’re looking for in an AI model, beyond maybe saying they want the best one.

Rather than establishing AI selection criteria, we choose instead to defer our choice of AI to what we see or hear in the news.

The AI media ecosystem has conditioned us to think about model selection the way we think about sports rankings. Who’s number one? Which model topped the latest evaluation suite? The implicit assumption is that “best at benchmark tests” translates to “best for your specific use case,” even though that assumption is almost always wrong.

Models like Google’s Gemma 4 and DeepSeek v4 have been attracting serious attention in the AI engineering community, not because they top every leaderboard, but because they deliver comparable performance to much larger frontier models at a fraction of the infrastructure cost, a sign that the battleground for AI model supremacy is moving to the frontier of AI efficiency over pure AI accuracy.

DeepSeek v4 is priced at roughly one-sixth the cost of leading competitors, generating what the developer community is calling a “second DeepSeek moment,” a nod to the market impact that DeepSeek had when it was originally unveiled to the world in 2025.

What models like DeepSeek v4 and Gemma 4 are demonstrating is that it’s now possible to get near-frontier performance at a fraction of the cost, and that the gap between the “best model” and “the right model for the job” is widening. Developers are already prioritizing the use of cost-efficient AI models alongside tools like Claude Code for complex coding tasks instead of strictly relying on the thinking models that top the LLM leaderboards.

They’re making strategic decisions at the task level that most senior leaders in organizations haven’t started thinking about yet. Developers are embodying the financial discipline to match the right model to the right task at the lowest cost that still meets the performance bar.

This way of thinking about AI is showing up elsewhere. Just have a look at how China has weathered GPU export bans that were intended to constrain its own AI development. Instead of inhibiting AI, it’s forced an efficiency pivot that’s become a structural competitive advantage. The constraint has become the edge.

Fit isn’t the same as performance

AI model selection is more than just a niche engineering consideration. It’s actually a leadership and commercial one, and one that many leaders have yet to make deliberately.

It’s really a decision of fit, rather than one of performance: match the right model to the right task, deliberately, based on what conditions for success matter most to your organization instead of just relying on external rankings.

A model that scores highest on a reasoning benchmark may respond in a style that doesn’t work for your team’s workflows. A model that’s technically less capable may follow complex, multi-part instructions more reliably, which matters far more in practice when deploying real-world AI workflows than marginal differences in raw intelligence.

Also: a model that communicates in a way that matches your organization’s culture will be adopted more readily than one that doesn’t, regardless of what the benchmarks say.

We might be tempted to think about “non-performance” considerations the same way we think about “hard skills” versus “soft skills” when we evaluate job candidates, but it’s more than that. These variables of fit are the ones that determine whether an organization will actually use AI, or shuffle it to the side after the initial enthusiasm fades.

To know what AI model really is “the best,” leaders need to be asking what models are the best fit for the work being asked of AI, in the context of the people who will be using it.

Six things to audition for

The good news is that evaluating fit doesn’t require a technical background. It requires the same skills you already use when hiring: structured observation of how a candidate performs on the things that actually matter for the role.

Here are six things you can do today to evaluate AI models against any application you’re considering. Look for:

1. Reasoning transparency. Ask the model to walk you through how it would approach a decision where evidence points in two different directions. The conclusion doesn’t matter. It’s whether the model shows its work. A model that externalizes its reasoning is easier to audit, challenge, and trust.

2. Instruction precision. Give the model a multi-part instruction with a specific format requirement and an unusual constraint. Count how many parts it fulfills completely and correctly. That consistency is one of the most important practical variables for complex workflows and automations, often more important than raw intelligence.

3. Cultural adaptability. Ask the model to rewrite an executive summary for a skeptical board audience, then rewrite it again for a frontline technical team. How does the model adapt its communication style? Do tone, vocabulary, and structure change? Or is it just producing two versions of the same document? Some models are better than others in adapting to different work environments.

4. Intellectual honesty. Ask the model for the strongest arguments against a position you’ve just asked it to defend. Spar with it in a live debate to see how it responds. A model that engages seriously with counterarguments, including arguments that challenge its own previous output, is more valuable than one that optimizes for agreement.

5. Internal consistency. Run the same complex reasoning prompt twice, separated by time. Compare the structural approach, not just the content. Meaningful variance in how a model reasons through the same problem is a signal worth paying attention to before you embed it in a critical workflow.

6. Cost efficiency. Compare the inference costs of the various models you’re considering. Unit costs are often represented in the form of input and output tokens, but some providers use virtual credits that correlate to token use. Once you’ve identified candidates that pass the five dimensions above, cost efficiency becomes a tiebreaker: what model gets you the results you need at the lowest cost?

I want to flag that there are a range of LLM parameters (you might run into terms like temperature, Top-P, and Top-K) that one can tune to customize how an LLM responds to a prompt and will influence how an AI model performs against these six criteria. Some of these are variables that are only accessible in a development environment or through API access to a model, so be aware that model selection is potentially even more nuanced.

Score each model against these six dimensions, weighted by what matters most for your specific use case. A team using AI primarily for written communication will emphasize factors like cultural adaptability more heavily than a team embedding AI into complex automated workflows, which crucially depend on instruction precision and internal consistency above all else.

The right model (or the right model with the right choice of parameters) is the one that performs best on the dimensions that bear the load of the work you’re asking AI to do.

Choosing the right AI model (or AI models – plural) for your organization is a consequential decision. Not because there’s competition to see who ranks first on a leaderboard, but because choosing well is a decision that directly impacts the financial and operational performance of your organization.

Access to AI models is no longer the constraint. Choice is. Hugging Face, one of the leading repositories for AI models, shares that its community has published over two million AI models for others to use. Organizations are literally spoiled for choice.

Instead of looking at how to get access to the “most powerful” models, there’s value in developing the management discipline to evaluate fit deliberately, deploying AI with precision, and resisting the pull of both emotional attachment and market-driven rankings.

While AI might not be a colleague in the same way a person can be, it deserves a proper interview before you give it the job.

AI, Upfront publishes every Monday. If this was useful, subscribe to get it in your inbox. And if there’s a topic you’d like me to tackle, reply and let me know.

Subscribe now

I've been writing AI, Upfront for the better part of a year. Thirty-plus briefings. One briefing each week.

If you're arriving for the first time, I won't ask you to read from the beginning. The archive is long and not everything in it reflects where this newsletter is heading. Instead, here are three pieces I'd point any new reader to first. They represent the thinking at the core of what I write about, and they'll tell you quickly whether this is worth your time.

If you want to understand why most AI initiatives fail before they start: Briefing #1: The Great AI Stagnation

Most AI projects don't fail because the technology doesn't work. They fail because organizations design pilots that are built to succeed on the wrong terms: technically sound, strategically hollow. If you've ever watched a promising AI initiative die quietly in a steering committee, this one is for you.

If you want to understand what AI fluency actually requires of a leader: Briefing #4: From AI Literacy to AI Fluency

There's a difference between knowing how to use an AI tool and knowing how to think with one. Most corporate AI training stops at the first. This piece makes that distinction sharply and explains why closing the gap between the two is the single most important thing a leader can do right now.

If you need to make the business case for AI to a skeptical CFO: Briefing #30: How to Talk to Your CFO About AI

This one came directly from a conversation with a senior tech executive whose AI initiative was dead on arrival, not because the idea was bad, but because he was telling the wrong story to the wrong audience. I've been in that room on both sides of the table. This piece walks through how to reframe an AI investment in language that actually moves a CFO from skeptic to ally.

From here, new briefings publish every Monday. Starting now, you’ll see longer pieces, video, and a sharper focus on what leaders in your position can actually do with AI — not just think about.

If there’s a challenge you’re navigating or a topic you’d like me to address, reply to this post. I read everything.

— Barry

Note: This briefing was originally published on LinkedIn on March 20, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There is a ghost that haunts every boardroom and every strategic offsite: the ghost of initiatives past. It’s the collection of ambitious plans and three-ring binders that were launched with fanfare, only to be silently archived months later, leaving the business entirely unchanged. We’ve spent decades perfecting the art of the pilot project and the “skunkworks” team, yet an astonishing number of these efforts still today fail to create lasting value.

As we stand at the precipice of the AI era, we, as leaders, are at risk of repeating the same mistake on a more expensive and accelerated scale. We can feel the pressure to “do something” about AI, but the path forward often seems like a dense, impenetrable fog.

If there’s anything to take away from the briefings I’ve shared with you over the past several months, it’s that the only way out is through. The companies who will win in today’s AI climate aren’t going to be the ones with the most visionary, long-range plan. They won’t be the ones with the best AI models, or even the most compelling AI-powered use case.

Those organizations that succeed will be the ones who have mastered the art of the sprint. The ones who build a high “organizational metabolism” that lets them consistently turn small, focused wins into compounding institutional knowledge.

They understand that the aim of any transformation effort goes beyond the launch of a perfect AI system, but the development of an internal engine for change that endures long past any particular transformation initiative.

This is what I mean when I say this is “the end of the beginning.” The work to mastering AI isn’t about learning to “do AI.” It’s about learning to think like an AI-native leader – to see your business not as a collection of static processes to be optimized, but as a dynamic system of capabilities waiting to be re-imagined, augmented by AI.

It might seem like the path forward to helping your business become AI-native is mythical and mystical. I’d suggest something totally different: it’s no mystery. It’s a discipline:

• Diagnose your most critical business challenges

• Build human-centric systems that address them

• And codify your success so the entire organization gets smarter.

Then, do it again. That transformation flywheel you’re creating – watch where it takes you when it begins to turn.

Note: This briefing was originally published on LinkedIn on March 20, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

For years, we’ve been told that data is the new oil, especially in an AI climate where data fuels the very AI models that generate enormous capability.

It turns out this is only half true. Raw data is indeed a commodity, but the most precious aspect of it is derived as a by-product of data creation: it’s the proprietary knowledge your team generates when they turn that data into a successful business outcome that’s most valuable.

It’s the “why” behind the what – the hard-won wisdom that comes from navigating your organization’s unique challenges and learning from how such wisdom can be deployed to the organization’s benefit over time.

And most companies let it vanish into thin air.

I have worked with a number of aerospace companies over the course of my career, and all of them were deeply concerned with the loss of institutional knowledge. What would happen to their engineering prowess when their most senior engineers departed the firm? How would the organization best ensure younger engineers keep that knowledge alive in succeeding them?

All enduring organizations are subject to that challenge. When an employee leaves, along with it goes the institutional memory that this individual held. The same is true for every AI initiative, pilot or otherwise. What happens when the initiative is marked as “complete” by the team that led it? How would other teams and other parts of the organization know what was learned?

Without a system for capturing and sharing what was learned, the institutional memory gained from every AI project gets lost. Many organizations are capable of doing excellent work, but they may be poor at systematically learning from it.

This is often the default state for most businesses. These companies operate with a slow “organizational metabolism,” burning immense energy on individual projects but failing to convert that effort into lasting institutional strength.

In the AI-native era, a slow metabolism is a death sentence. But there is a remedy: building a lasting “learning organization,” where the goal of an AI strategy is not to complete a series of disconnected AI projects, but to develop an ingrained organizational capability that is a dedicated engine for change.

It’s an organization that recognizes its most valuable asset is the ever-growing AI playbook of what it has learned. It has a formal process for “locking in” the insights from every success and failure, ensuring that the entire organization gets smarter with each initiative.

This is what transforms a one-off AI project into a compounding institutional asset, and it’s the single greatest source of durable competitive advantage where every business is “doing AI.”

Note: This briefing was originally published on LinkedIn on March 13, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There’s an old parable about a master cartographer. He spent his life creating the most detailed map of the empire ever known. It was a perfect 1:1 scale replica, so vast and so precise that it captured every road, every river, every single building. The map was a marvel of data collection. It was also completely useless. To unfold it was to cover the very empire it was meant to describe.

We are now entering the age of the 1:1 map.

With the power of AI, we can generate an almost infinite amount of data, analysis, and insights. We can have a 100-page report on a new market in seconds. We can have a real-time dashboard tracking a thousand different metrics. We have more “answers” at our fingertips than ever before.

And yet, many leaders could be feeling more lost than ever. They’re drowning in seas of data, experiencing a kind of analysis paralysis. The map has become the territory, instead of the guide it was intended to be.

Many organizations look at the act of data analysis as the impetus for “data science” or “AI” teams with remits to review reams of data and apply it to their organizations. This might come from the tacit belief that the challenge we have with using data to make better decisions comes from our inability to access or process information.

Here’s a thought: what if, instead of standing up “data teams,” everyone in an organization could be a “data sense-maker?”

As AI improves in its ability to pour through vast amounts of data, making sense of data and knowing where to attend to it remains a challenge. “What does the data mean?” is a question that will be familiar to any leader who’s been in an analytics presentation.

The opportunity is there for leaders to provide what AI can’t: to provide meaning to all the analysis being done by AI.

This signals a fundamental shift in the economics of leadership. For decades, a leader’s value was in the ability to make a decision based on experience and incomplete data. Now, as AI provides more and more of the data-driven “doing,” the leader’s value is shifting to the strategic “sensing.”

Strategic sense-making is the uniquely human ability to:

• Find the Signal in the Noise: To look at a thousand data points and identify the one that truly matters.

• Interpret Complexity: To understand that data can tell you what is happening, but it can’t tell you why.

• Create a Narrative: To weave the data into a coherent and compelling story about what to do next.

A junior analyst can use AI to summarize a report. A leader must use their judgment to determine what that summary means for the business.

This “sense-making” is human value and uniquely human context not amenable to automation. As answers become a commodity, the ability to ask the right questions and create a clear point of view may very well be the rarest and most valuable leadership skill of all.

Note: This briefing was originally published on LinkedIn on March 6, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

For the past two years, the dominant metaphor for AI in the enterprise has been AI as “co-pilot,” a comforting and brilliantly marketed concept. It suggests a helpful assistant, sitting at your side, ready to take over routine tasks and ease your day-to-day.

But I believe this metaphor, while well-intentioned, brings along with it an implicit bias that limits the way in which we think about the role that AI can play in our work. It encourages a master-servant relationship with technology.

We delegate our thinking, we passively accept the AI’s first answer, and we optimize for comfort and efficiency. This, as I’ve shared in the past, is a direct path to cognitive dependency and organizational fragility.

A far more powerful, albeit more challenging, model is to think of your AI not as a co-pilot, but as a sparring partner.

A co-pilot’s job is to make you comfortable. A sparring partner’s job is to make you better. A sparring partner challenges you, tests your defenses, and exposes your weaknesses in a safe environment, before you step into the real ring with a competitor or a customer.

This positions AI not just as an interface for delegation, but as a partner through which constructive dialogue is possible.

Instead of asking the AI, “What are the top 5 markets for us to enter?”, a leader in a “sparring partner” culture would say, “Our top market is Germany. Now, act as our biggest competitor and tell me why this is a terrible idea.” The goal of this interaction is not to get a faster answer. It’s to build a more robust strategy by subjecting it to rigorous, critical debate.

This is the “Socratic Dialogue” in a modern context. It’s an iterative process of questioning, challenging, and refining that uses AI’s immense knowledge base and capacity for analysis to sharpen our own human judgment.

Cultivating this culture requires a new kind of leadership. It means explicitly rewarding the employee who challenges the AI’s output, not the one who accepts it fastest. It means celebrating “productive friction” as a sign of a healthy, intelligent team. It means training your people not just on “how to prompt,” but on “how to question.”

The co-pilot model promises a smoother ride. The sparring partner model promises a stronger team. Choose wisely: do you want to build a business that’s merely efficient, or one that makes the best decisions?

Note: This briefing was originally published on LinkedIn on February 27, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

I recently spent time with a senior tech executive at an industrial B2B company who was deeply frustrated. His team had built an impressive AI prototype that would have fundamentally changed his organization’s ability to address inbound customer inquiries. Instead of serving customers on service loops that typically require hours, it might be possible to serve customers in minutes. But his business case to take the AI prototype further was dead on arrival.

“The CFO just doesn’t get it,” he said. “He keeps asking for a hard ROI, and nothing I share seems to land well.”

Sadly, this is a story that’s more common than maybe it should be. It’s a clash of two valid, but different, views of the world. On the one side, there is the innovator who sees a strategic opportunity. On the other, there is the steward of the balance sheet who sees an unquantified risk.

The problem isn’t the obsession with ROI, or what some might conclude as the debate between the quantitative versus qualitative aspects of operating a business. The problem is in the translation of innovation to measurable outcome.

In our rush to champion the transformative potential of AI, we often forget that “transformation” is not a line item on a balance sheet. AI-native leaders who are successful at gaining the alignment they need to do “big” things are more than visionaries. They’re translators – they’ve learned how to deconstruct their ambition into a financial story that a CFO (and Board) can analyze, measure, and ultimately invest in.

Arguably, the most obvious form of ROI is what is easily measurable: cost savings. As in, “This AI will automate X dollars of hourly labor.” But ROI can take other forms – and it’s in being able to tell the ROI story from multiple angles that turns a business case for AI from a mere transaction to a strategic endeavor.

There are at least three distinct kinds of value that can form the basis of a sophisticated presentation of the financial return on an AI investment:

1. Cost Efficiency: This is the classic ROI of cost reduction and automation. It’s the simplest, most direct part of such a business case.

2. Revenue Growth: This is AI’s potential to generate net-new revenue. It’s not just about making your existing business cheaper, but about using a new AI capability to create a new product, service, or business model. This is where you might quantify the size of a new market you can now address, or the new revenues that might emerge from an entirely new revenue stream.

3. Enterprise Value: This is the most abstract, but also the most durable form of value. This is the proprietary data and insight an AI bespoke to your organization that leads to long-term value from competitive or market differentiation and the strategic moat that might set your organization apart from others. This is a balance sheet asset capable of generating returns for years to come.

By separating the business case into these three parts, you can have a more honest and complete conversation by showing how AI creates near-term value, and enables a path to long-term, strategic value. Instead of asking for a budget, you’re presenting an investment portfolio.

Note: This briefing was originally published on LinkedIn on February 20, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

Imagine driving a high-performance car. You’re obsessed with the dashboard. The tachometer is screaming at 8,000 RPM, the speedometer says you’re doing 120 miles per hour, and the engine temperature is perfect. By all these metrics, your vehicle is operating at peak performance.

But then you look up, and you realize you’ve been driving in a perfect circle for an hour. You haven’t actually gotten any closer to your destination.

This is the state of AI measurement in many enterprises today.

We’re building beautiful dashboards that glow with green vanity metrics. We celebrate when our new chatbot handles 10,000 “conversations” in its first month. We high-five when our sales team “adopts” the new AI-powered CRM feature. It turns out we’ve been measuring the engine’s RPM instead of our time to destination.

The implications of chasing poor metrics are profound. Poorly-conceived metrics allow us to invest millions of dollars and thousands of hours into initiatives that may make us “feel good,” albeit only for a moment, while delivering zero business value.

While it isn’t wrong for an organization to be proud of having achieved a 90%+ adoption rate of a new AI capability, it must always be assessed in the context of the P&L. Perhaps uptake has been swift, but did it actually move the needle? Perhaps the team did play with the tool – once, but hasn’t touched it since. Such an AI dashboard would be green while the P&L remains red.

The antidote is to shift our focus from activity to capability.

AI-native leaders don’t ask, “Is the AI busy?” They ask, “Is our business getting better because of the AI?”

This requires a new type of metric. Instead of measuring the tool, you must measure the business capability the tool is supposed to improve.

• Don’t measure “chatbot queries handled.” Measure “customer retention rate” or “cost-to-serve.”

• Don’t measure “AI tool adoption.” Measure “sales cycle time” or “lead conversion rate.”

These are Capability Metrics. They’re the only metrics your board, your CFO, and your shareholders actually care about. It’s the only way to know if you’re actually moving forward, or just driving in circles with the engine screaming.

Note: This briefing was originally published on LinkedIn on February 13, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

For the past year, the conversation around AI governance has been dominated by a narrative of fear. We talk about risk mitigation, compliance checklists, and the potential for brand-damaging errors from “hallucinating AI” or AI “leaking” data to the outside world. We have, with good intentions, framed governance as a necessary evil — a safety brake we must apply to AI to ensure we deploy it responsibly.

This might profoundly undersell the opportunity that well-conceived AI governance creates.

While your competitors are viewing governance as a burdensome cost, a new class of AI-native leaders is in the process of making it the unique selling point of their AI-enhanced products and services. They understand a fundamental truth of this new era: as AI becomes more powerful and autonomous, the customer’s decision-making process is no longer just about the features you offer. It’s about the trust you inspire.

In this environment, the way in which you enact governance becomes more than just a set of mechanics to ensure compliance. It’s a product feature. In fact, it might be your most important one.

Consider the landscape:

• Access to Regulated Markets: In sectors like finance and healthcare, the barrier to entry isn’t technology, it’s regulatory approval. A company that can demonstrably prove its AI is safe, transparent, and fair through a robust, platform-based governance model can enter these lucrative markets faster and more effectively than a competitor with a “move fast and break things” ethos.

• Winning the Enterprise Customer: The CIO of a Fortune 500 company is not just buying your AI tool. They are buying your assurance that it will not expose them to a billion-dollar lawsuit or a front-page data breach. When two vendors offer similar features, the one with the superior, more transparent governance model will win the enterprise deal every time.

• Building Brand Equity: In a world awash with “AI slop” and concerns about data privacy, a brand that can verifiably claim “our AI is trustworthy” has a powerful and rare differentiator. Trust is becoming a premium attribute, and customers will pay for it.

The most mature organizations adopting AI are doing more than simply asking “How do we get this past legal?” They’re also asking “How do we feature our superior guardrails in our next sales pitch?” These organizations are turning their SOC 2 reports and their real-time monitoring dashboards from internal compliance documents into external marketing assets.

This is a call to stop thinking of governance as a tax on innovation. Start thinking of it as the foundation for your next premium product.

Your strongest shield can also be your sharpest sword.

Note: This briefing was originally published on LinkedIn on February 6, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

For the past decade, one of the most powerful concepts in business strategy has been the “Data Flywheel.” Pioneered by companies like Amazon, the logic was simple and powerful: a better product attracts more users, who generate more data, which is used to make the product even better. The company with the most data wins.

That era is over.

In the age of generative AI, the value of raw data is collapsing. The Internet is being flooded with low-quality, unreliable, AI-generated “slop.” AI models trained on this contaminated data are beginning to degrade, a process researchers call “model collapse.”

Simply having “more” data is no longer an advantage. In fact, it can be a liability if the quality is poor. The strategic high ground has shifted. The new moat isn’t built on having the most data. It’s being built on having the best process for creating proprietary data.

The last generation of “Big Data”and the “Data Flywheel” is giving rise to a new, more powerful model: the Judgment Flywheel.

If the old Data Flywheel was a giant vacuum cleaner, sucking up all available data, the Judgment Flywheel is a refinery. It takes crude, generic inputs (like an AI’s first-pass analysis) and uses a deliberate, human-centric process to refine them into high-grade, high-signal intelligence.

Here’s how it works:

1. AI Output: An AI agent makes a first-pass decision (e.g., flags a transaction). This is the crude input.

2. Human Judgment: A human expert reviews the decision and, if they override it, provides the critical “why.” This is the refining process.

3. Captured Insight: That “why” — the expert’s judgment — is captured as a structured, proprietary data point. This is the high-grade fuel.

4. Smarter AI: This new, high-quality data is fed back into your AI models, making them smarter and more contextually aware than any generic system.

The cycle repeats, with each turn making your entire system more intelligent and more aligned with the unique realities of your business. This is what creates a true, unassailable competitive advantage. The winner is no longer the company with the biggest data lake, but the one with the most effective intelligence refinery.

Note: This briefing was originally published on LinkedIn on January 30, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

In the world of structural engineering, there’s a critical difference between strength and resilience. You could build a bridge out of a material that is incredibly strong but perfectly rigid. It would stand flawlessly for years, right up until the day a unique, high-frequency vibration — from an unusual wind pattern or a specific traffic load — hits its resonant frequency.

The bridge would shatter catastrophically. A resilient bridge, by contrast, is designed with built-in flex and dampening systems, allowing it to bend without breaking.

Many organizations today are building “brittle bridges.”

In our quest for AI-driven efficiency, we are architecting “perfect” processes that handle 99% of cases flawlessly. We automate the invoicing system to process an exact data format. We build a chatbot to answer a specific list of questions. But this pursuit of perfection creates extreme fragility. The system works, until it doesn’t.

A supplier adds a single new field to their invoice. A customer asks a novel question. Or, a small innocuous typo appears in an input table. The whole workflow grinds to a halt.

Contrary to what we might think, this isn’t a failure of technology. It’s a failure of imagination. We’re applying an industrial-era, assembly-line mindset to a dynamic, digital world.

An AI-native organization operates with a different mindset. Instead of building processes, they design systems. A process is rigid; it breaks on exceptions. A system is adaptive; it learns from exceptions.

I saw this firsthand in my time in financial services. Banks often feel compelled to ensure nothing ever fails. And in so doing, they apply major constraints on customer interactions. To verify your identity correctly, you must provide the exact full name you used when you got your account. You must know the exact date a transaction occurred. Or, heaven forbid, there’s an Internet outage – you might have to enter your carefully provided input all over again.

A more resilient approach might consider inputs as probabilities, rather than absolute truths. “Unexpected” inputs get flagged as exceptions, but they don’t kill the transaction. They might get routed to a person – or an AI sub-agent – to see if there’s a means to disambiguate what’s been provided. But they don’t penalize the user because he might have forgotten to include his middle name.

We all know from our experience that few things in life are certain. If we’re looking to build AI workflows that move the needle, the goal isn’t to build AI that never fails. It’s to build systems that never stop improving.

Note: This briefing was originally published on LinkedIn on January 23, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

I recently had a conversation with a senior executive who was reminiscing about his early career, a time before desktop computers were ubiquitous. He confessed a nagging fear. “I worry,” he said, “that we’re losing our knowledge of first principles.”

“With AI, will we be able to solve problems on our own ever again?”

That’s a question I’ve pondered since. It’s timely to be sure – we see a version of this playing out in headlines about the “AI cheating epidemic” in colleges and universities, where course instructors are becoming increasingly reliant on oral exams because their confidence is waning around students’ willingness to do assignments and written tests unaided.

As one undergraduate student recently wrote in a powerful plea in Maclean’s Magazine, a culture of “copying, pasting, rephrasing and submitting” is creating an environment where software talks to software, while learning becomes an afterthought.

Other publications like the Harvard Business Review are talking about “cognitive offloading,” or the reliance of skilled talent at work on AI to do tasks they used to do just fine themselves, leading to burnout – or worse.

For enterprise leaders, the well-intentioned desire to show measurable results from our AI investments is a risk vector for creating this dynamic. We are building a generation of “AI-literate” employees who are exceptionally good at prompting an AI, but we might be failing to cultivate “AI-fluent” teams who know how and when to apply sound judgment to its output.

Cognitive offloading in an organization is the atrophy of the critical thinking, problem-solving, and first-principles reasoning that an organization needs to survive a true crisis. In the desire to become “AI-native” (or “AI-first,” pick your term), we’re creating teams that are poised to be incredibly efficient at scaling problems AI already knows how to solve, but that become dangerously fragile when faced with a novel challenge that requires a human to step up.

The antidote is not to reject AI. It’s to be more deliberate and strategic in how we deploy it. We must stop thinking of “AI-native” as a buzzword for “deploying AI everywhere.” A more practical and resilient definition is: building an organization around the unique and complementary skills of humans and AI.

How can leaders act on this today? Here are two practical places to start:

1. Design “Graceful Workflows.” Instead of building workflows that are fully dependent on AI, design them to be “graceful.” A graceful workflow is one that can be powerfully augmented by AI, but that could still be fully executed by your human team if push came to shove — if the cloud goes down or the model starts to drift. This builds resilience directly into your operations.

2. Trial with Humans First. Before you hand a critical workflow over to an AI agent, have your most trusted human experts run the process manually for a week (or more). This does two things: it ensures the workflow and customer experience actually make sense, and it allows a designated “red team” of your best people to vet the process and identify the real risks before automation scales them.

This measured approach ensures that AI serves the business, not the other way around. It allows us to gain the efficiencies of AI without sacrificing the human judgment that ensures our long-term resilience.

Note: This briefing was originally published on LinkedIn on January 16, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

For the better part of a century, the blueprint to becoming a successful leader was clear: you excelled at managing people. From the industrial-era supervisor ensuring tasks were completed to the corporate-era manager facilitating team success, the core function was always human-centric.

With AI on every leader’s mind, the end of that era of leadership could well be at hand.

AI systems are now capable of handling the very tasks that once defined management: tracking performance, monitoring workflows, and optimizing output. This automation of oversight doesn’t necessarily make leaders obsolete though. It liberates them to do more important work.

The challenge is that this new work requires a radically different skill set.

The old model was about getting things done through others. The new model that’s emerging is about achieving outcomes by orchestrating a complex ecosystem that includes both human talent and “AI talent.”

Leaders in this AI era are more like symphony conductors. Conductors don’t play a single instrument, but they have the ability to understand the unique capabilities of every section – the strings, the woodwinds, the percussion – orchestrating them to produce transcendent music that no single section could create alone.

Your human team might be your string section, unmatched in creativity, empathy, and strategic judgment. Your AI agents might be your percussion, capable of immense scale, speed, and pattern recognition. Your job is to write the score that allows them to play in harmony and manage the full orchestra as it performs live.

Successfully navigating this new landscape requires mastering a new set of core competencies:

1. Human-AI Collaboration Design: Blending the strengths of humans and AI.

2. Agent Management: Prompting, evaluating, and coordinating AI outputs.

3. Strategic Sense-Making: Interpreting complexity and maintaining focus amidst a sea of AI-generated data.

Leaders who insist on defining their value by the number of people they manage will quickly realize AI brings with it not just one, two, or ten direct reports – but the challenge of simultaneously managing dozens or hundreds of team contributors.

At that scale, the very notion of what it means to provide leadership must evolve. It’s in this way that the future belongs to leaders who see themselves as orchestrators of outcome.

Note: This briefing was originally published on LinkedIn on January 9, 2026. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There’s little doubt 2025 was the year of the AI initiative. Enterprise businesses are estimated to have spent between $400 billion to an eye-watering $1.5 trillion on AI in the past 12 months alone.

That’s an unfathomable amount of investment — especially when just 6% of organizations report experiencing “significant” value from their AI investments.

What are the 94% who aren’t experiencing significant returns from AI doing?

They’re funding prediction models, building chatbots, and “AI-powered” features, declaring victory as soon as they can show off a demo. When these demos fail to garner support or the organizational momentum needed to carry on, they’re quietly shelved — or replaced by the next shiny object. The business itself remains unchanged.

The historical precedent for this age of AI we’re in isn’t the Internet, the smartphone, or even the desktop PC. It’s actually something many of us likely take for granted today: it’s electricity.

Let me explain: the companies that won the electric age weren’t the ones who simply bought dynamos to light up their existing, inefficient factories. The winners were the ones who realized the new technology allowed them to fundamentally redesign the factory itself — creating the modern assembly line and unlocking massive gains in productivity.

They didn’t just use electricity. Instead, they built their entire operating model around it.

This is actually the AI-native mindset.

It wasn’t too long ago that many businesses were building “innovation” or “digital” teams, proudly appointing Chief Innovation Officers or Chief Digital Officers with a mandate and deep pockets to lead sweeping organizational transformations. I spent time with one such organization some years ago, an insurance provider, who had made impressive investments to secure a leadership position in the early days of AI, even before the current wave of generative AI we’re all in today.

They had an innovation team including outstanding PhDs in AI and data science. They also had a strong portfolio of AI initiatives. But, the team was disconnected from those who owned P&L. They became a “cost center,” arms length from the actual business. Ultimately, they proved unable to generate discernable value — and were quietly disbanded.

They were a company with AI, but they were not an AI company.

A company that embraces the AI-native mindset looks at AI very differently. They view their core business processes as AI products themselves. A shipping company’s product isn’t “moving boxes,” but rather, is a dynamic engine for “logistics optimization,” for example. These businesses don’t look at AI like it’s a technology you buy, but more like AI is an integral part of the very infrastructure that enables the business to operate.

That 6% of organizations who are seeing significant growth in their EBIT from AI also happen to be the ones actively seeking the AI-native mindset. They’re the ones redesigning processes and workflows, targeting areas where they can scale quickly, and making investments to transform the organization, not just their tech stacks.

Adopting an AI-native mindset means making a fundamental shift in how you invest in AI, who you hire, and where you create value. The call to action is surprisingly simple.

Stop funding demos. Start rewiring your factory.

Note: This briefing was originally published on LinkedIn on December 19, 2025. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There’s a cynicism gripping the market, and it’s not hard to understand why.

For years, the narrative of business has been dominated by a singular focus: delivering shareholder returns by extracting maximum value for minimal effort. This has led to a consumer landscape where it’s easy to feel constantly nickeled and dimed, sold products that feel less durable, and served by systems that prioritize cost-cutting over customer experience.

Now, into this environment, we’ve introduced AI. It maybe isn’t a surprise then that in some corners of the market, the reaction to AI has been one of skepticism.

When companies deploy AI to automate an already frustrating customer service experience, consumers react accordingly. They ask the question: “What exactly am I paying for?” They may see AI not as an innovation, but as the next logical step in a long history of companies trying to do as little as possible.

This is today’s great, unspoken tension. And it’s why the mainstay playbook of business could well be on the cusp of implosion.

I’m reminded of a lesson from early in my career in financial services. A wise senior leader shared with me that heavily regulated industries, such as banking, are sometimes slow to change because those in charge have become stewards of a system they grew up with. Such leaders have a hard time envisioning a world where the system they were taught to be “the way things are done” might one day no longer exist.

That same dynamic isn’t unique to banking. It’s playing out across every industry today. Leaders are clinging to an outdated playbook, not because it’s effective, but because it’s familiar.

But a new playbook is being written. Its central tenet is not efficiency, but value creation.

The old playbook asked: “How can we optimize this process?”

The new playbook asks: “What new form of value can we create that makes the old process obsolete?”

This is the strategic shift that AI enables. AI is not just a tool for optimization; it’s an engine for imagination. It offers us the chance to partner with a non-human intelligence, a creative collaborator that can help us reframe problems we once thought impossible to solve. It allows us to offload the work we are merely competent at, freeing us to flourish in the areas where we are truly brilliant.

As we head into 2026, the question is not whether we will face challenges — we will. The question is what direction we’ll choose to pursue.

Will we continue to defend pre-AI ways of thinking that are becoming increasingly irrelevant? Or will we embrace the disruption, carve out new ways of doing things, and, in the process, discover new ways to define success?

The future that awaits us will be decided by the clash of these two playbooks. Will we believe the world can never be any different? Or will we believe that the future ahead is what we make of it?

Our only limitations will be the ones we place on ourselves.

Note: This briefing was originally published on LinkedIn on December 12, 2025. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There’s an old joke in computer science that the way to solve any complex problem is simply to abstract it and make it the next layer’s problem.

For decades, the corporate world has applied this same logic to organizational design. When a business becomes too complex for one leader to manage, we “abstract” the problem by hiring a layer of managers. When those managers get overwhelmed, we hire directors to manage them. We solve the complexity of scale by building a human hierarchy.

It works, but at a steep cost. We trade agility for control. We create a “frozen middle,” a thick layer of translation where strategy gets diluted and ground truth gets distorted before it ever reaches the C-suite.

But today, in this era of AI we’re in, we’re seeing the signals of a profound reversal.

In October, Amazon announced it was cutting 14,000 corporate roles in a bid to build flatter hierarchies across its corporate organization and invest more heavily in AI. Life sciences giant Bayer AG was also recently in the news touting its operating model, one of “Dynamic Shared Ownership,” where up to 95% of the organization’s decision making has been pushed to the front lines in a trend that some are calling the “Great Flattening.”

This is organizational thinking that goes beyond layoffs or efficiency. It is a statement of belief in the potential for AI to enable an architectural upgrade of the corporation itself.

The New Abstraction Layer

What’s obvious to a computer scientist isn’t always obvious in business: the goal of abstraction isn’t to add more people. It’s to hide complexity so you can focus on the result.

In the past, “abstracting complexity” meant hiring a middle manager to summarize weekly reports from ten junior employees. The manager was the API.

Today, agentic AI is becoming that abstraction layer.

Instead of relying on a deep organizational structure to filter information, leaders can now insert AI into their business models to observe results and take action directly.

• The Way It’s Done Today: A CEO waits for a quarterly report, synthesized by three layers of management, to know if a product launch is working.

• The “Founder Mode” Way: An AI agent monitors customer sentiment, sales velocity, and support tickets in real-time, flagging anomalies directly to leadership the moment they happen.

This achieves more than the flattening of an organization. It compresses the distance between the leader’s intent and the organization’s action. It allows a $500 million enterprise to operate with the visceral, hands-on agility of a $5 million startup.

The Small Business Giant

This shift frames the future of work we are heading toward. The “AI-native” organization of 2026 will look less like the sluggish incumbents of the Fortune 500 and could look more like a massive network of small, agile businesses.

These organizations will be leaner, yes. But more importantly, they will be “sludge-free.” They will use AI not just to automate tasks, but to automate the coordination of tasks — the very thing that middle management was invented to do.

For leaders, this is the ultimate AI unlock. It means you may no longer have to choose between scale and speed. You could have the reach of a giant with the soul of a founder.

Note: This briefing was originally published on LinkedIn on December 5, 2025. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

I once had a customer, a senior leader on the cusp of retirement, who told me, “I’m sure our company must change and embrace new technologies. But not while I’m still here and have anything to say about it.”

It was a bold, candid admission. It’s also a perfect illustration of the perilous, personality-driven environment in which many organizations operate, where real transformation opportunities often go to die.

Today, we see this kind of human friction creating a new stalemate in the face of organizations seeking to wrangle and reason about AI. A polarized, all-or-nothing debate around AI has emerged, trapping organizations in a cycle of inaction. This isn’t an “AI divide” born from technology, but a leadership divide born from hubris.

On one side, we have “AI shamers.” These are the folks who proudly point out the perceived tics of AI-written content, as if spotting an em-dash or a certain prosaic pattern makes them a guardian of authenticity. They’re also the ones who fear most the decline of work quality through reliance on AI as a crutch.

On the other, we have “AI evangelists” who insist AI must immediately take over all “low-value” work, a dangerously subjective term. Who, exactly, gets to decide what’s “low-value?” They’re the ones who fear most about becoming irrelevant and being left behind.

This binary thinking — this “my way or the highway” mindset — is a failure mode. It’s the same impulse that drives leaders to tear down the ideas of others just to assert their own authority. Much has been said in the past year about the importance of executive alignment in driving successful AI adoption, and yet, many organizations struggle to achieve said alignment because debates around AI can be so charged they turn into sparring matches over who’s right and who’s wrong instead.

This is where the real villain of AI transformation is exposed: the people-driven impulse that values being right over getting it right.

Pragmatism teaches us neither polar position is correct. Leadership wisdom suggests resisting a simple “yes” or “no” to AI gets us to wrangle a more difficult, nuanced answer: “It depends.”

In this context, “it depends” isn’t an act of avoidance but the start of a strategic diagnosis about what role AI plays inside an organization, and where and how it can be employed to achieve the best outcomes. And equally important, what areas are off-limits to AI:

• It depends on the problem you’re solving.

• It depends on the risk of the workflow.

• It depends on the data you have.

• It depends on the business outcome you need.

For example, using AI to co-author a low-risk internal memo is a completely different strategic decision than deploying an agent to autonomously handle a high-risk financial compliance workflow.

One outcome values speed and “good enough” efficiency while the other demands 100% accuracy, where a single error or hallucination could be catastrophic. Without asking when and why, leaders are just blindly adopting technology (or just as blindly, refusing to consider it), rather than strategically deploying it.

The most valuable leaders in the next decade won’t be the ones who can spot an AI-generated email. They will be the ones who fight hard for the alignment that requires teams, leaders, and organizations to get past hubris and force a dialogue that confronts the real issues at hand.

The answer to “Should we use AI for...?” isn’t “yes” or “no.” It’s “Here’s when.”

Note: This briefing was originally published on LinkedIn on November 28, 2025. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

There’s a philosophical debate raging right now, sparked by articles like one I recently read in the New Yorker: Is AI really thinking, or is it just a cleverly complex mimic?

It’s a captivating question. For those of us leading a business, maybe the real question isn’t “What is AI?” but “What do we believe it is?”

How your organization answers this question or, more likely, how it avoids answering it could define your entire approach to AI. Over the past year, I’ve observed that the business world has fractured into at least three distinct camps, each defined by its own implicit point of view on what AI’s role in business is:

1. The Status-Quo Champions: These are leaders who claim to champion AI but are really only interested in maintaining the status quo. They use AI as “innovation theater.” They’ll fund a small, contained chatbot pilot that goes nowhere. For this group, AI isn’t a tool. It’s a topic for a press release. Their unstated belief is that AI is a mild force, one that can be managed and contained within an existing way of working, broken processes and all.

2. The Exploiters: These are the people seeking to exploit AI because they think it’s a path to a quick dollar. They see AI as a crude automation tool, a way to cut costs and replace headcount. They aren’t interested in transformation, only in extraction. Their unstated belief is that AI is a blunt instrument, and they’re all too happy to wield it, often damaging customer trust and employee morale in the process.

3. The AGI Worshippers: These are the people waiting for divine inspiration. They believe AI’s true homecoming will arrive with the emergence of Artificial General Intelligence (AGI). Until that day of rapture, they argue, it’s all just “wait and see.” Their unstated belief is that AI is a spectator sport. They are paralyzed, waiting for a future event rather than acting in the present.

The one thing all three camps have in common? They’re all, in their own way, passively waiting to be told what AI is. They’re letting the hype, the fear, or inertia dictate their strategy.

This is a failure of leadership. To paraphrase Cisco’s John Chambers: “Disruption waits for no one.” While it may feel prudent to wait and see how things play out, leaders must be asking these questions now, not later in response to a disruption that’s already taken place.

So, what’s the alternative?

The alternative is to decide. The best leaders aren’t waiting to be told what AI is. They’re actively deciding what role AI will play in their organization and building a strategy to match.

Here are three things you can do today to position AI effectively in your organization and evergreen your business for the disruptions we don’t even know about yet:

• Open Your Eyes: Encourage experimentation. Assume the next disruption could happen tomorrow. The goal isn’t to find a single, perfect use case. The goal is to build your team’s “AI fluency.” Give them the permission and the safety to explore, to test, and to fail small so they can learn fast.

• Tune Off the Hype: Come to your own conclusions. Ground your strategy in the reality of where your organization is today and what ultimate value it could be creating. The market-at-large doesn’t know your business, your customers, or your team. Stop listening to the hype that tells you what AI “should” be, and start defining what it will be for you and your organization.

• Be the Disruptor: Don’t wait for change to come so you can react to it. Take the initiative and forge the path that lets your organization be the disruption, rather than the one being disrupted. Identify your own most valuable, complex process and ask, “What if we were the ones to make this obsolete?”

We started this article by asking what AI really is. Here’s the bottom line: AI is what we make of it.

We can either be told by others what AI is, or we can, as leaders, decide. The choice is ours.

Note: This briefing was originally published on LinkedIn on November 21, 2025. It has been migrated to our new home on Substack to create a complete archive. Multi-format features like video and audio commentary are available for all new briefings published from April 2026 onwards.

In the face of massive technological change, every business leader is eventually forced to answer one very unexpected question: Is your organization a Nokia or a Nintendo?

Nokia is, arguably, the quintessential case study in complacency. Once so dominant that “Nokia” was synonymous with “mobile phone,” the company’s roots were actually in 19th-century paper and rubber goods. They were, for a time, masters of reinvention. But at the peak of their success, they fell into a trap: they dismissed the first iPhone not as a revolutionary ecosystem, but as a niche, expensive toy.

This isn’t just a historical case study for me. I saw this firsthand during my time there.

From the inside, one would have seen that there wasn’t a single moment of failure, but a continuous cultural shift that played out every day. Nokia had become trapped by its own success. Internal teams focused on optimizing what the organization believed was most important to consumers: hardware durability, call clarity, and battery life. The internal narrative, reinforced in meeting after meeting, was that the iPhone’s “flaws” — a fragile glass screen, poor battery, and no physical keyboard — made it a non-starter for “serious” users.

Nokia was executing a perfect plan for a world that was about to disappear. Data and past successes didn’t so much blind the organization rather than cause it to focus on the wrong things. At times, that’s what complacency feels like: a failure of imagination, borne from the comfort of success. The company was disrupted because it stood still.

Then there’s Nintendo. Their story is not one of flawless, linear victory. They also started in the 19th century, making playing cards. They dominated the 80s and 90s, but their history since has been a mix of blunders (like the Virtual Boy) and industry-defining successes (like the Nintendo Switch). Earlier this year, Nintendo launched the Switch 2, which claimed the record of being the fastest selling home video game console of all time.

The key difference is this: Nintendo never stopped trying to reinvent. They embody imperfect, relentless action.

This distinction is the entire lesson for leaders navigating the age of AI. The anxiety so many feel doesn’t come from the technology itself. It comes from the organizational paralysis in the face of it. It’s the fear of being a Nokia.

The prevailing winds today suggest AI may excel at taking exams but struggles with basic arithmetic. It’s tempting to see these limitations as a shield, a reason to wait and see, just as Nokia waited to see if the smartphone vision of the iPhone would bear out. But history shows this is a fallacy. Relying on today’s limitations to predict tomorrow’s capabilities is how incumbents get blindsided.

The antidote isn’t a perfect, multi-year strategic plan. It lands on the wisdom that “a good plan executed today is better than a perfect plan next week,” to paraphrase World War II General George Patton.

Execution, however imperfect, is the only antidote to disruption.

Pragmatic leaders who internalize this share three core habits. They exemplify the disciplined, daily work of fighting complacency.

1. They Actively Reject the Status Quo. “It’s the way we’ve always done things” is a warning sign. These leaders choose, instead, to embody the spirit of a George Bernard Shaw quote, most famously used by Robert F. Kennedy: “Some men see things as they are and ask, ‘Why?’ I dream of things that never were and ask, ‘Why not?’“

2. They Never Rest on Their Laurels. Intel’s legendary co-founder and CEO, Andy Grove, built his entire leadership philosophy on this idea, famously captured in the title of his book: “Only the Paranoid Survive.” This isn’t about anxious fear, but rather the relentless, productive vigilance against the inertia that success breeds.

3. They Master the “Art of the Possible.” Bill Gates’ vision of a “PC on every desk” seemed absurd at the time. It didn’t happen overnight. It was the result of pairing a huge ambition with “eminently doable” next steps, executed relentlessly. This is the very definition of pragmatic innovation.

The conversation today is about AI. Tomorrow, as the famed futurist Ray Kurzweil notes in what he calls the “Law of Accelerating Returns,” the next disruption will likely emerge even faster, and from a direction we don’t expect.

There is one memorable footnote worth sharing about Nokia. In a powerful testament to this very idea of reinvention, the Nokia of today has found new life, pivoting entirely from mobile phones to become a leader in network solutions. This itself is the point: reinvention, even after a catastrophic failure, is always possible.

The ultimate takeaway for pragmatic leaders is not to predict the future. It is not to wait for the “perfect” future. Instead, “it is,” as the visionary computer scientist Alan Kay once famously said, “to invent it.”