Almost everyone arguing about artificial intelligence and jobs is arguing about a different thing, and mistaking it for the same thing. Spend a week with the commentary and you will meet a dozen confident, incompatible verdicts. One says a white-collar bloodbath has already begun. Another says technology always creates more work than it destroys, so relax. A third says AI only changes tasks, not jobs. A fourth says the machines will make us all so rich that work becomes optional. They cannot all be right, and the temptation is to pick a side and defend it. I spent the past several weeks doing something different. Working with a small salon of human and AI minds, set to challenge one another rather than to agree, I sorted the debate into ten recognizable camps and put each one to a simple test: what would have to be true for this position to stand, and what observation would prove it wrong? Then I checked each against the evidence as it actually stood in mid-2026. The result surprised me less for who turned out to be wrong than for why everyone was talking past everyone. The camps do not disagree about the facts nearly as much as they appear to. They disagree about which danger matters most. Here is what the data actually saw. Across the advanced economies, unemployment sat at 4.9 percent, squarely inside the band it has held since 2022. The forecast of immediate mass joblessness is, so far, simply wrong. Yet beneath that calm surface, entry-level job postings had fallen by more than a third, and by more than forty percent in the most AI-exposed roles, while young workers in those roles lost around thirteen percent of their employment relative to their older colleagues. The aggregate is steady and the bottom rung is breaking at the same time. That is not a contradiction. It is the signature of what I have come to call the Net Employment Chasm, the gap that opens when tasks are destroyed quickly and new ones are created slowly. Jobs return, but not on the clock they leave, and the people who fall into the gap are disproportionately the young, trying to gain a first foothold. Two further findings complete the picture, and both are confirmed. The newest economic modeling shows that AI can narrow the gap in wages while it widens the gap in wealth, because the returns to the machines flow to the concentrated few who own them. And a state that funds itself by taxing labor faces an eroding base at exactly the moment displacement raises the demand for public support. So we hold three distinct risks, and they can all be true at once: a broken entry rung, a widening concentration of wealth, and a hollowing tax base. A society can reach full employment and still end up more unequal and less solvent. That is why the camps talk past one another. The Bloodbath forecaster fears the lost job. The Distributionist fears the captured gain. The Fiscal realist fears the empty treasury. Each is right about a different danger, and answering one does nothing to settle the others. Once you see the debate this way, the useful question stops being who wins the forecast and becomes which of these risks you are carrying, and with what instrument. A word on the most dangerous financial advice of the year. You have likely seen the claim, offered from the very top of the industry, that you no longer need to save for retirement because AI will deliver an age of abundance and a universal high income. I would treat that counsel with great care. Cheaper production is not the same as distributed access. A promise that you will be provided for is not a transfer of ownership, and in all of economic history abundance has never once distributed itself. It was distributed, when it was, through institutions that had to be built, and only because the powerful still needed the many as workers and as customers. Notice, too, that the people most confident money is about to become irrelevant are the ones accumulating capital, compute, and land at record speed. Their revealed preference is louder than their forecast. Which brings me to the part I care about most. I am not a doomer. I see in these systems an immense possibility for human flourishing, provided we build and use them within an ethical frame. The gap between what AI removes and what it restores is real and structural, and it will not close on its own. But a bridge across it is buildable, and a costed plan to build one can be judged rather than merely admired. To make that concrete, I subjected one such plan, from my own book, The Great Reimagining, to an independent simulation of two hundred thousand scenarios, and I published where it holds and where it breaks. Its base case is affordable, its crisis provisions prove necessary rather than decorative, and its one real vulnerability is its dependence on how large AI's productivity gains actually turn out to be. I would rather tell you that honestly than sell you a soft landing. That is the shift I am arguing for. Stop trying to win the forecast, and start treating AI and work as a portfolio of distinct risks, each with its own instrument. Protect the first rung, because the entry roles being automated are the stepping stones through which people once climbed to the senior work that AI cannot yet touch. Broaden ownership rather than tax it heavily, so the returns to the machines reach the many rather than the few. And rebuild the tax base deliberately, before the erosion forces the question under worse conditions. Plans of this kind should be judged on their arithmetic, neither waved away as advocacy nor swallowed as gospel. Judging them is how the conversation moves from warning to building, which is where, on the evidence, it now needs to go. The full working paper, "The Structural Employment Chasm and the Buildable Bridge," lays out all ten camps, the test each must pass, and the simulation in detail. I would genuinely welcome your disagreement. A claim earns its standing by surviving the attempt to refute it, and this one is offered in exactly that spirit. Copyright © 2026 by Severin Sorensen. All rights reserved.

Nearly seven in ten C-suite executives report being seriously close to leaving their roles for reasons tied to their well-being (Silverglate & Fisher, 2022). That survey predates the current wave of AI adoption, but a more recent study gives the pattern a name and a mechanism. Researchers at Boston Consulting Group surveyed nearly 1,500 full-time workers and found that intensive AI oversight, rather than AI use itself, is driving a distinct form of mental fatigue they call "AI brain fry," marked by difficulty focusing, slower decision-making, and higher turnover intentions (Bedard et al., 2026). In other words, the exhaustion isn't coming from delegating more to AI. It's coming from what AI adoption asks leaders to personally hold, verify, and reconcile in real time. Workload vs. Cognitive Load Workload is a volume problem. You address it by delegating, automating, or saying no. Cognitive load is a coordination problem. It comes from holding multiple unresolved, often contradictory demands in mind at once and having to keep making sound decisions anyway. In 2026, AI adoption has become one of the primary sources of that load for the leaders responsible for it, which creates a strange irony: the technology meant to relieve pressure on the organization is intensifying pressure on the people steering its adoption. Consider what a typical executive is now expected to hold simultaneously. The board wants a credible AI strategy and a return on the capital already committed to it. Employees want reassurance that adoption will not cost them their jobs, delivered in language specific enough to be believed. Customers expect the polish AI promises without any visible seams. Regulators are still writing the rules the leader is meant to already be following. None of these demands resolves on its own timeline, and a leader cannot simply postpone one to focus on another. They coexist, and the executive is the only place where they all have to be reconciled at once. This is decision fatigue in its more dangerous form. It is not the familiar tiredness of a long day of meetings. It is the erosion of judgment that comes from constant context-switching between fundamentally different value systems, financial return, workforce trust, technical feasibility, and public accountability, all in the same afternoon, often in the same conversation. The BCG/HBR research on AI brain fry found this pattern is strongest in exactly this kind of high-oversight role, where marketing and HR functions reported it most acutely (Bedard et al., 2026). In coaching conversations, these shifts tend to show up in behavior well before they show up in a performance review. A leader with a strong instinct for people starts making calls that seem uncharacteristically flat. Meetings get shorter, not because the leader has grown efficient, but because their bandwidth for nuance has quietly narrowed. A Design Problem, Not a Resilience One Most organizations misdiagnose this as a personal wellness problem and respond accordingly, with a meditation app, an executive health screening, or a well-intentioned reminder to take a vacation. Those responses are not wrong so much as they are aimed at the wrong layer. They treat the leader as the site of the problem, when the leader is more often the last stop for a problem the system generated upstream. A wellness app cannot resolve the fact that a CEO is being asked to have a defensible point of view on AI governance, workforce redesign, vendor selection, and competitive strategy, all before most of the relevant standards and best practices have had time to mature. The more useful frame, and the one that should inform how coaches and organizations approach this moment, is that AI-era burnout at the top is substantially a design problem before it is a resilience problem. Leaders are not failing to cope. They are being asked to serve as the integration layer for decisions that no single person, however capable, is well positioned to make alone and in real time. The fix starts with reducing how much of that integration work has to happen inside one person's head. Three Shifts to Reduce Cognitive Load Three shifts help, and none of them require a leader to become more resilient in the abstract sense the word usually implies. Sequence Decisions The first is sequencing decisions rather than holding them open simultaneously. Many executives treat every AI-related question as urgent and current, which means they are perpetually revisiting the same four or five unresolved threads without ever closing one. A structured decision calendar, where governance questions, workforce questions, and vendor questions each get a dedicated window rather than competing for attention in every conversation, does more to reduce cognitive load than any amount of stress management technique. It converts an undifferentiated pile of pressure into a sequence, and sequences are something the mind can actually process. Distribute Judgement The second is distributing judgment rather than centralizing it further. A recurring pattern among executives navigating AI adoption is a reflexive instinct to personally validate more decisions, not fewer, because the stakes feel higher and the technology is unfamiliar. This instinct is understandable and almost always counterproductive. It concentrates load exactly where it should be relieved. Coaches working with these leaders should be asking directly which decisions genuinely require the executive's judgment and which have simply defaulted there out of habit or anxiety. Reduce Coordination Burden The third is using AI itself to reduce the leader's coordination burden rather than adding to it. Much of what currently lands on an executive's desk, synthesizing conflicting stakeholder input, tracking the status of parallel workstreams, drafting the fifth version of a communication to employees, is exactly the kind of structured, well-bounded work that AI tools handle competently. Used this way, AI functions as a release valve on cognitive load rather than another demand on it. The distinction is not about which tools a leader adopts. It is about whether adoption is designed to lighten the leader's coordination burden or simply adds a new category of decision to the pile they are already carrying. The Main Takeaway None of this argues that leaders don't need genuine rest, support, or attention to their own wellbeing. They do, and that need is real. But treating burnout purely as a personal deficit lets organizations off the hook for the structural conditions actually producing it, and it leaves leaders trying to out-willpower a problem that was never about willpower to begin with. The leaders who come through this period well will not be the ones who found a way to tolerate more. They will be the ones who, with the right support, restructured how much they were being asked to hold at once. That is a coaching conversation, a governance conversation, and increasingly, a conversation about how AI itself gets deployed inside the leader's own workflow, not just the organization's. References Bedard, J., Kropp, M., Hsu, M., Karaman, O. T., Hawes, J., & Kellerman, G. R. (2026, March 5). When using AI leads to "brain fry." Harvard Business Review. https://hbr.org/2026/03/when-using-ai-leads-to-brain-fry Silverglate, P. H., & Fisher, J. (2022, June 22). The C-suite's role in well-being: How health-savvy executives can go beyond workplace wellness to workplace well-being—for themselves and their people. Deloitte Insights. https://www.deloitte.com/us/en/insights/topics/leadership/employee-wellness-in-the-corporate-workplace.html Copyright © 2026 by Severin Sorensen. All rights reserved.

Most executives treat exercise and diet as variables they will optimize later, once the quarter closes or the deal clears. A major clinical trial published last year suggests that the “later” instinct carries a measurable cost, and that the fix is far more specific than the general advice to “eat well and move more” that fills most wellness content aimed at leaders. The study is the U.S. POINTER trial, a two-year, five-site randomized controlled trial run by the Alzheimer’s Association and published in JAMA in the summer of 2025 (Baker et al., 2025). Researchers enrolled 2,111 adults between the ages of 60 and 79, all considered at elevated risk for cognitive decline due to factors such as sedentary habits, a suboptimal diet, cardiometabolic risk, or family history of memory impairment (Baker et al., 2025; Alzheimer’s Association, 2025). Participants were split into two groups. One followed a self-guided approach, receiving general encouragement and health education but choosing their own path. The other followed a structured protocol built on four pillars, exercise, nutrition, cognitive training, and health monitoring, with prescribed goals, frequent check-ins, and accountability built into the design (Baker et al., 2025; Baker et al., 2024). The structured group’s prescription was specific on both fronts. On exercise, it called for roughly 30 to 35 minutes of moderate-to-intense aerobic activity four days a week, plus 15 to 20 minutes of resistance training twice a week and 10 to 15 minutes of flexibility work twice a week (Baker et al., 2024; “Lifestyle Medicine and Brain Health,” 2025). On nutrition, participants received ongoing coaching to follow the MIND diet, a hybrid of the Mediterranean and DASH diets built specifically around brain health, emphasizing dark leafy greens, berries, nuts, whole grains, olive oil, and fish, while limiting red meat, fried food, sweets, and added salt (Alzheimer’s Association, n.d.). That combination, layered with computerized cognitive training and regular monitoring of blood pressure, weight, and lab results, produced measurably better outcomes than the self-guided approach (Baker et al., 2025). The structured group showed greater improvement in global cognitive function, and researchers estimated the intervention protected participants from up to two years of the cognitive decline typically expected with normal aging (Baker et al., 2025; Alzheimer's Association, n.d.). The benefit held consistent across age, sex, and genetic risk subgroups (Baker et al., 2025). What makes this trial worth an executive’s attention is that exercise and diet matter for long-term health. What is new is the evidence that structure itself, not just the underlying behaviors, is the active ingredient. Both groups in the trial exercised more and ate somewhat better than at baseline. Both groups received encouragement. The group that outperformed the other did so because their program had defined goals, a fixed cadence, and built-in accountability across both domains at once, the same discipline that separates a functioning executive team from a well-intentioned one (Baker et al., 2025). For a population of leaders who already run their calendars against quarterly targets and their teams against KPIs, this finding reframes brain health as a governance problem rather than a willpower problem. The same discipline applied to a P&L or a board deck, translating vague intent into specific, trackable targets, appears to be what separates measurable cognitive protection from good intentions that quietly erode under travel schedules and back-to-back calls. There is also a strategic argument buried in the data. Executives are, by definition, in the business of long-horizon decision making, and cognitive endurance is now a documented input to that capacity, not an assumed one. A board member or founder in their sixties or seventies who follows a structured protocol, on both the training and nutrition side, is, according to this trial, meaningfully better positioned to sustain the judgment, memory, and processing speed that the role demands (Alzheimer’s Association, 2025). Building the Physical Training Protocol The trial’s structured arm delivered exercise through coach-supported sessions at community facilities, resources most executives will not replicate exactly (Baker et al., 2024). But the underlying architecture, specific weekly targets across cardio, strength, and flexibility, paired with a mechanism for accountability, can be approximated using AI as a planning and tracking layer. Prompt 1: Build the baseline training protocol “Act as an exercise physiologist. Design a weekly training protocol for a [age]-year-old executive with a demanding travel schedule, modeled on the structure of the U.S. POINTER trial: roughly 30 to 35 minutes of moderate-to-intense aerobic activity four days a week, 15 to 20 minutes of resistance training twice a week, and 10 to 15 minutes of flexibility work twice a week. Account for [any injuries, equipment access, or time constraints]. Present it as a seven-day table with exercise type, duration, and intensity target for each day.” Prompt 2: Adapt for travel weeks “Take the protocol above and create a compressed, equipment-free version that fits into hotel rooms and airport layovers, preserving the same weekly ratio of four cardio sessions, two strength sessions, and two flexibility sessions, but capping each session at 25 minutes.” Building the Nutrition Protocol The trial’s nutrition arm was not a generic instruction to “eat healthier.” It was ongoing coaching toward the MIND diet, with specific weekly targets for food groups shown to support brain health (Alzheimer’s Association, n.d.; Baker et al., 2024). That same specificity can be built into a personal meal-planning system. Prompt 3: Build the baseline MIND diet plan “Act as a registered dietitian. Build a one-week MIND diet meal plan for a busy executive, based on the nutrition prescription used in the U.S. POINTER trial: daily servings of dark leafy greens and other vegetables, regular servings of berries and nuts, whole grains over refined grains, olive oil as the primary fat, fish at least once a week, and minimal red meat, fried food, sweets, and added salt. Account for [dietary restrictions, cooking time available, or meals eaten out]. Present it as a seven-day table with breakfast, lunch, dinner, and one snack.” Prompt 4: Adapt for travel and restaurant meals “Take the MIND diet plan above and translate it into a restaurant-ordering guide: what to look for and what to avoid on a typical hotel breakfast menu, an airport terminal, and a business dinner, while staying close to the MIND diet targets for leafy greens, whole grains, olive oil, and fish.” Building the Accountability Layer The trial’s real advantage was not either domain in isolation. It was tracking both together against a fixed cadence (Baker et al., 2025). Prompt 5: Build a combined weekly check-in “Design a simple weekly self-check-in template, modeled on the accountability structure used in clinical lifestyle trials, that tracks adherence to both a four-days-cardio, two-days-strength, two-days-flexibility training protocol and a MIND diet nutrition plan. Include a short weekly reflection prompt and a way to flag when adherence to either domain drops below 75 percent for two consecutive weeks.” The Main Takeaway None of these prompts require specialized software, and none depend on wearable data or a nutritionist on retainer, though either can sharpen the output further. What they require is the same discipline executives already apply elsewhere: defined targets across both domains, a fixed cadence, and a way to notice when adherence slips. The U.S. POINTER trial suggests that discipline, applied to the body and the plate for a few hours of planning a week, pays a dividend in the very capacity leaders depend on most, the ability to think clearly and decide well, for longer. References Alzheimer’s Association. (n.d.). U.S. POINTER study results. Retrieved July 2026, from https://www.alz.org/us-pointer/study-results Alzheimer’s Association. (2025). U.S. POINTER shows structured lifestyle program targeting multiple risk factors improves cognition in older adults at risk of cognitive decline. Alzheimer’s & Dementia. https://doi.org/10.1002/alz.70608 Baker, L. D., Espeland, M. A., Whitmer, R. A., Snyder, H. M., Leng, X., Lovato, L., Papp, K. V., Yu, M., Kivipelto, M., Alexander, A. S., Antkowiak, S., Cleveland, M., Day, C., Elbein, R., Tomaszewski Farias, S., Felton, D., Garcia, K. R., Gitelman, D. R., Graef, S., Howard, M., … Carrillo, M. C. (2025). Structured vs self-guided multidomain lifestyle interventions for global cognitive function: The US POINTER randomized clinical trial. JAMA, 334(8), 681–691. https://doi.org/10.1001/jama.2025.12923 Baker, L. D., Snyder, H. M., Espeland, M. A., Whitmer, R. A., Kivipelto, M., & U.S. POINTER Study Group. (2024). Study design and methods: U.S. study to protect brain health through lifestyle intervention to reduce risk (U.S. POINTER). Alzheimer’s & Dementia, 20(11), 8140–8147. https://doi.org/10.1002/alz.13365 Lifestyle medicine and brain health: Insights from the U.S. Study to Protect Brain Health Through Lifestyle Intervention to Reduce Risk (U.S. POINTER) and the promise of personalization. (2025). [Journal article]. PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC12624361/ Copyright © 2026 by Severin Sorensen. All rights reserved.

That which is old is new again. The thought has crossed my mind repeatedly as I have watched generative AI models emerge, improve, and regenerate over the past several years, and it has returned with particular force as I write the third edition of The AI Whisperer. Stephen Covey's counsel to begin with the end in mind, offered to business readers in 1989, turns out to be precisely what is needed in this new age of specification. The humble terminal window, brought to life more than half a century ago, persists today as the gateway to Claude Code and a growing family of agentic tools. The pattern resembles a Fibonacci sequence, in which each new term is the sum of the terms that came before it, and the growth that appears sudden is in fact the accumulation of everything preceding it. This observation carries a practical lesson. The most powerful advances in artificial intelligence do not discard history. They revive, scale, and recombine proven ideas with new compute and new data, and organizations that recognize these continuities adopt AI more effectively than those chasing only what feels novel, because they can link new tools to strengths they already possess. What follows are ten patterns in which an older idea has returned at the center of modern AI. I begin with the most relatable and proceed toward the more deeply technical, and I close with the strongest objection to my own argument, because a thesis untested is a thesis unearned. 1. Goal Specification: Begin with the End in Mind Covey's second habit holds that effective action starts with a clear mental picture of the desired result. In the current era of AI, which some are calling AI 3.0, this principle has become an operating requirement rather than a motivational aphorism. Precise prompts, agent task definitions, and explicit success criteria are the mechanisms through which human intent becomes machine output, and the quality of what a model produces tracks closely with the clarity of the specification it receives. A vague request yields generic material, while a request that carries a vivid vision of the outcome, the audience, and the standard of customer delight yields work a leader can actually use. Covey wrote The 7 Habits of Highly Effective People in 1989 for managers organizing human effort, and his principle now governs how we organize machine effort as well. I place it first not because it is the oldest idea on this list but because it is the most immediately actionable one, and because it opens the door to everything that follows. 2. The Terminal and the Command Line The text-based terminal descends from the earliest days of interactive computing, with Unix arriving at Bell Labs in 1969 and the POSIX standards formalizing its conventions in 1988. One might have expected graphical interfaces to retire it decades ago. Instead, AI coding agents and automation systems have made the command line newly central, because it offers precision, scriptability, and deterministic execution that visual tools struggle to match. An agent that can read and write plain text in a shell can operate almost any system ever built, which is why the oldest interface in computing has become the preferred hands of the newest intelligence. 3. Neural Networks and Connectionism No item on this list embodies the thesis more completely. Warren McCulloch and Walter Pitts proposed a logical model of neural activity in 1943, Frank Rosenblatt introduced the trainable Perceptron in 1958, and the field was then largely set aside after Marvin Minsky and Seymour Papert exposed the limits of single-layer networks in 1969. Backpropagation, popularized through the 1986 work of David Rumelhart, Geoffrey Hinton, and Ronald Williams, revived the approach, and the arrival of large datasets and graphics processors allowed it to triumph at scale. The idea at the heart of every large language model today is an idea that was proposed, doubted, abandoned, and vindicated across eight decades. 4. Reinforcement Learning and Feedback The principle that behavior improves when actions are followed by rewards or corrections traces to Edward Thorndike's law of effect in 1911 and to the operant conditioning research of B. F. Skinner in the mid-twentieth century. Richard Sutton and Andrew Barto formalized the computational framework, and reinforcement learning from human feedback now shapes the alignment and preference tuning of frontier models. A century-old finding from animal psychology has become the finishing school of artificial minds. 5. Conversational Agents The experience of conversing naturally with a machine feels unprecedented at today's scale, yet the underlying loop is old. Joseph Weizenbaum's ELIZA, built at MIT in 1966, used simple pattern matching to sustain dialogue, and Weizenbaum was unsettled by how readily users attributed understanding to it. Modern systems have replaced his handwritten rules with learned representations and vast context windows, but the essential structure of turn-taking, and the human tendency to meet the machine halfway, remain exactly as he found them. 6. Retrieval-Augmented Generation Allowing a model to consult external sources before answering reduces error and keeps knowledge current without retraining. The retrieval half of this marriage is venerable. Hans Peter Luhn explored statistical term weighting in the late 1950s, Gerard Salton developed the vector space model of retrieval through the SMART project in the 1960s and 1970s, and Karen Spärck Jones introduced inverse document frequency in 1972. Patrick Lewis and colleagues formalized retrieval-augmented generation in 2020, joining sixty years of information retrieval research to modern generation. 7. Self-Supervised Learning Large models acquire their capabilities by predicting missing or subsequent elements of raw data, without armies of human labelers. The lineage runs through the unsupervised learning and autoencoder research of the 1980s, which pursued the same goal of extracting structure from unlabeled data but lacked the scale to demonstrate its full power. The idea waited forty years for its data and its compute to arrive. 8. Mixture of Experts In 1991, Robert Jacobs, Michael Jordan, Steven Nowlan, and Geoffrey Hinton proposed adaptive mixtures of local experts, an architecture in which specialized subnetworks handle different portions of a problem and a gating mechanism routes each input to the appropriate specialist. The idea remained a curiosity for a quarter century until Noam Shazeer and colleagues demonstrated sparsely gated mixtures at scale in 2017, and variants of the architecture now sit inside several frontier models. Few examples illustrate the return of the old more cleanly than a 1991 design serving as the skeleton of 2026 systems. 9. The Distributional Hypothesis and Embeddings Every embedding, and therefore every semantic search and every retrieval pipeline, rests on a claim made by linguists seventy years ago. Zellig Harris argued in 1954 that words occurring in similar contexts carry similar meanings, and J. R. Firth compressed the insight in 1957 into a maxim that deserves to be quoted in full: "You shall know a word by the company it keeps." Word2vec operationalized the hypothesis at scale in 2013, and the transformer's representation of meaning as position in a learned vector space is the same idea carried to its logical conclusion. Hold on to Firth's sentence as you read what remains, because it turns out to describe more than words. 10. Symbolic Reasoning and Structured Task Design The expert systems of the 1970s and 1980s, exemplified by MYCIN and DENDRAL, captured domain knowledge in explicit rules and were eventually eclipsed by statistical learning. Yet their spirit has returned inside modern agent systems, which gain reliability when neural fluency is combined with explicit tools, verification steps, and structured plans. The same is true of task analysis, the mid-century discipline from human factors and systems engineering that mapped complex work into steps, goals, and edge cases before automating it. Effective prompt and agent design today is task analysis under a new name. The Strongest Objection: Sutton's Bitter Lesson An honest account must confront the counterargument. Richard Sutton argued in his 2019 essay The Bitter Lesson that seventy years of AI research teach one thing above all, namely that general methods leveraging computation ultimately defeat approaches built on human-crafted knowledge and structure. On this reading, history is a record of clever human ideas being swept aside by scale, and reverence for old patterns is precisely the error to avoid. The objection has force, but it cuts less deeply than it first appears. The methods that scale so triumphantly are themselves old ideas. Connectionism, reinforcement learning, and self-supervised prediction are the very lineages traced above, and what compute swept aside was not history but a particular kind of hand-engineering. Compute did not replace the old ideas; the old ideas were waiting for the compute. The bitter lesson and the present argument are therefore compatible. Scale determines which of the old ideas win, and the winners have so far been the oldest and most general of them. For practitioners the reconciliation is even simpler, because the durable human disciplines on this list, clear goals, reliable interfaces, feedback, retrieval, and structured task design, operate at the layer of applying AI rather than building it, and no amount of compute relieves a leader of the duty to specify what a system should accomplish. What This Means for Leaders The pattern across all ten examples is consistent. Progress in AI has come from taking durable ideas, removing their old limits with new scale, and combining them in fresh ways, which produces systems that feel revolutionary while resting on solid foundations. The practical counsel follows naturally. Leaders and teams gain less from chasing each new model release than from connecting AI to the proven methods already present in their organizations, including clear goal specification, dependable interfaces, feedback loops, structured decomposition of work, and disciplined use of external knowledge. That approach yields steadier adoption, lower risk, and stronger results, because it treats AI not as a rupture with the past but as the past returning with new power. Firth taught that you shall know a word by the company it keeps, and the same may be said of ideas. Each of the ideas gathered here carried one meaning in its own era, when its company was punch cards, laboratory rats, or a lone terminal humming in a Bell Labs basement. Placed in new company, among vast datasets, abundant compute, and one another, the same ideas have come to mean something larger, and that shift in company, rather than any rupture with the past, is what we are living through. The best AI strategies do not ignore history. They build upon it, with clarity and with purpose, knowing that ideas, like words, are made new by the company they keep. Copyright © 2026 by Severin Sorensen. All rights reserved.

Optimizing AI Token Spend, while Minimizing Expense This past weekend I conducted an experiment on myself as much as on a machine. Anthropic's new Claude Fable model had been available to me for three days, and the announcement had already come down that access would soon move out of standard subscription allowances and onto metered usage credits today. The clock was running. Rather than ration my remaining allocation cautiously, I decided to do the opposite. I resolved to spend it deliberately, completely, and on the highest-value work I could assemble, treating every token as precious fuel and flying the aircraft until the gauge read empty. This is a refined form of tokenmaxxing: not wasting tokens as fast as you can, but the opposite. How to invest them in highest-order projects as intelligently as you can, on a fixed budget of tokens. It's an adaptation of an old Goethe saying: "As in art, the best is good enough, on budget." I want to describe both the experiment and what it revealed, because the lessons extend well beyond one weekend with one model. They reach into how leaders should think about attention, scarcity, and the allocation of any resource that matters. The Setup: Choosing What Deserves the Best Anyone who has worked seriously with frontier AI models learns quickly that capability is not evenly needed across tasks. A great deal of daily work, such as summarizing a document, formatting a table, or drafting a routine email, can be done perfectly well by lesser models at a fraction of the cost. The scarce and expensive resource, whether measured in tokens, dollars, or rate limits, should be reserved for the work where the marginal intelligence of the best available model actually changes the outcome. So before I spent a single token, I asked the question every operator should ask of an expensive asset: what work is worthy of this? Or as Clayton Christensen might have said, what are today's most critical and crucial 'jobs to be done.' I settled on three substantial projects, each chosen because it demanded sustained reasoning over many hours, synthesis across large bodies of material, and creative extension beyond what I had done before. The first was curriculum development turning a past AI workshop into an online AI course, a full program of 78 video lessons that required coherent instructional design across the entire arc of the course, not merely a stack of disconnected scripts. The second project was a new book project, which meant reviewing my past manuscripts, my serial publications, new research, and roughly two years of accumulated musings, asking the model to hold that entire corpus in mind while evaluating what the new work should become. The third was the most speculative and the most fun: pondering and writing code for the needs of 2030. The decade's end is closer than it feels, and I wanted to press the model for everything it could do now, drawing on my past work but pushing it, ever so gently, to stretch further. I was pleasantly surprised by what came back. Along the way I gave the model supporting roles as well. It played IT consultant to troubleshoot technical issues as they arose. It helped me develop concepts for teaching first principles of AI to executives. Like in my AI Workshops, I had several AI plates spinning at once, and I kept my system connected through the weekend so the work could continue throughout the night without me, compressing what would otherwise have been weeks of effort into a weekend. The Ceiling: What Happens When the Fuel Runs Out The experiment had a second purpose beyond the work itself. I wanted to see how quickly a heavy, sustained, multi-project workload would exhaust the Fable allocation on my Anthropic $ 200-per-month Max plan. The answer arrived before the weekend was completed: it maxed out with about 85% of the work completed. What happened next is the part worth dwelling on. When my access dropped back to the previous flagship model (Opus 4.8), the productivity of the whole operation fell noticeably, and not because the ceiling of raw capability had lowered. The most disruptive symptom was continuity. Like a light switch flipping off, the system forgot where it was in the process, losing the thread of work that had been proceeding smoothly for hours. Let me be fair to the older model: it remains excellent, and six months ago I would have described it as remarkable. But once you have worked at the new level with Fable, the previous one occupies second place in your mind, and you feel the difference in every exchange. It's like putting your 2nd string out on the playing field when you know your better players are benched by a penalty. Or another metaphor: losing access to Fable was like returning to a smaller monitor after a week on a large one, where nothing has failed, yet the working environment feels constrained in ways one notices instantly. I offer this less as a complaint than as an observation about the psychology and economics of capability upgrades. Each new tier of intelligence resets the baseline of what feels acceptable, and that reset has consequences for how we as managers budget, how we plan, and how we feel about the tools we thought we loved last month. The Deeper Lesson: Scarcity Is a Teacher Another lesson. What surprised me most about the weekend was the discipline that deliverable-focused tokenmaxxing forced upon me, namely the ruthless prioritization of what deserved the best model's attention, turned out to be a better use of my own judgment than almost anything else I did. We have a role with AI, and it serves up as acumen, prioritization, taste, and discernment of 'what's next.' When a resource is abundant, we spend it carelessly, as if it had no cost, and when it becomes scarce and expiring, we suddenly discover our inner economist. I found myself asking, for every task that crossed my mind, whether it was worthy of the premium model or belonged with a lesser one, and this understanding did more to clarify my actual priorities for AI builds than any planning exercise I have run this year. The parallel to leadership is direct. We prioritize customers because not all accounts warrant the same investment of our finest attention. We prioritize conversations, giving our best coaching hours to the team members where development will compound. And, most importantly, we prioritize time with the people who matter most, because the hours of a life are the original metered resource, allocated to us without a published rate card and without the option to purchase more. If I am willing to architect an entire weekend around extracting maximum value from a temporary allocation of machine intelligence, I ought to bring at least that much intentionality to how I allocate myself. The Economics: Building a Roster, Not Buying a Star Several observations surfaced for me on the same weekend as the announcement of new pricing policies, tiers, and API throughput arrangements for the new model, which sharpened the question from an experiment into a budgeting decision. I am not yet convinced that I will increase my spend. My monthly investment in AI across providers is already substantial, and the arrival of a more capable and more expensive tier does not automatically justify a larger budget. What it justifies is a reallocation. The right mental model, I have come to believe, is that of a sports team coach. A coach does not play the star in every minute of every game. Each player is deployed for the situations where their particular utility is highest, and a place on the bench reflects strategy rather than any slight to the player. Applied to AI, this means routing work deliberately. The frontier model gets the problems where its marginal intelligence changes the outcome: deep synthesis, long-horizon reasoning, creative extension of hard problems. Capable mid-tier models carry the substantial middle of the workload. Commodity models handle the routine. Other providers' offerings fill the roles where they hold genuine comparative advantage. The objective is not to own the most expensive roster in the league but to win games at a sustainable payroll. For enterprises, this discipline is arriving whether leaders are ready or not. As frontier access shifts toward metered consumption, token allocation becomes a management question rather than an IT line item, and the organizations that develop the habit of asking "what work is worthy of our best intelligence" will hold a quiet but compounding advantage over those that treat all AI usage as interchangeable. What I Would Tell a Fellow Executive If you are weighing how to approach this new generation of models, my weekend suggests a few practical conclusions. First, run your own tokenmaxxing experiment. Take a bounded window of access and spend it entirely on your highest-value problems. You will learn more about the model's true ceiling, and about your own priorities, than any benchmark or review can teach you. Second, prepare for the fallback. Whatever tier you operate on, there will be moments when you drop to a lesser model mid-process. Structure your work so that context is preserved outside the conversation, in documents and artifacts the next session can pick up, because continuity is where the productivity losses concentrate. Third, budget by allocation rather than by aspiration. The useful question is not whether the frontier model is worth the price in the abstract, but rather which specific work in your portfolio justifies frontier pricing, and how the remainder should be routed to the rest of the roster. Fourth, and finally, let the exercise instruct you beyond the technology. The same scarcity logic that governs tokens governs your calendar, your coaching hours, and your evenings at home. The machine taught me nothing this weekend that the wisest people in my life had not already tried to teach me. It simply presented the lesson with a usage meter attached, and there is something clarifying about watching a gauge fall in real time. The fuel always runs out, and the only question that remains at the end is whether it was spent on what mattered. I would be interested to hear how others are approaching this transition. Are you increasing spend for frontier access, reallocating across providers, or waiting to see how the pricing settles? The comments are open, and so am I. Copyright © 2026 by Severin Sorensen. All rights reserved.

Last week, the United States marked two hundred and fifty years since its founding, and it is worth remembering what actually provoked the founders. The colonists who boarded ships in Boston Harbor in December 1773 were not protesting a large tax. The duty on tea was three pence a pound, and the Tea Act had in fact made tea cheaper. What they refused was the arrangement itself: a small, metered charge on every pound, levied by a distant monopoly, with no voice in the terms, arriving on top of a decade of stamp duties, customs fees, and navigation restrictions. Small unit charges, compounding at scale, imposed without representation. That was one of many grievances worth a revolution. Artificial intelligence (AI) has brought us astonishing tools, and I have spent the last three years teaching business leaders how to use them. But the current AI economics have a familiar shape. Enterprises and API users now pay for intelligence by the token, a small metered charge on every unit of work, set by a handful of distant providers, on terms no customer negotiates, and at a scale where fractions of a cent compound into budget lines that boards are beginning to question. And these tokens are charged on throughput, regardless of whether the output is useful. The frustration is no longer confined to private conversation. On Wednesday, July 1, 2026, Palantir CEO Alex Karp told CNBC's Squawk Box that something in the token model "has gone completely wrong," and that "every single enterprise I deal with is livid," with customers telling him "I am paying for tokens that create no value." I hear the same sentiment, in nearly the same words, in one-on-one conversations with business executives at all enterprise levels. The public outrage and the private grumbling have converged. Some of the grievance, though, is self-inflicted, and this is the part we can control. Most people use AI sloppily, not from carelessness of character but because they never before needed to think about the cost of a word. The habit formed in an unmetered world, and it persists in a metered one. Field reports from operators who have audited their production stacks suggest that 40 to 60 percent of token budgets in production applications is pure waste: money paid for capability never used, or for inefficiencies nobody priced at the design stage. Here is the good news. If tokens are the measured medium, then the medium can be managed, and managed dramatically well. I will add one candid caution: it is possible that as enterprises grow efficient, providers will devise new ways of charging for what amounts to oxygen for AI agents. That prospect is an argument for the discipline, not against it, because the organizations that measure their consumption are the ones that notice when the terms change, and the ones with the standing to push back. Karp's remedy is to own the means of production. For most organizations, the nearer remedy is to govern the consumption. That is what my new book is about: identifying and stopping token waste, and implementing the procedures that turn AI spend into a strategic advantage. Today I published TokenOps: From Token Waste to Competitive Advantage, and its argument begins with a pattern we have seen before. The Pattern We Have Seen Before In 2012, the average enterprise cloud bill was a curiosity item on the desk of the chief financial officer. By 2018, it had become a top-five line in many technology budgets, and a discipline called FinOps had emerged to govern it. The companies that built FinOps competence early found themselves operating cloud infrastructure at materially lower unit costs than their less disciplined competitors, and the gap widened year on year. That sequence was not an accident of the cloud era. It is the characteristic pattern by which new operational disciplines come into being. A previously unmetered or flatly priced resource becomes commercially metered. Organizations adopt the resource faster than they can measure it. Costs accumulate silently and unevenly across teams. Executive attention eventually arrives, and a discipline emerges to bring order, supported by tooling, certification, and dedicated roles. Tokens are now traversing the same trajectory. They flow through every prompt, every retrieval-augmented query, every agentic workflow, and every multi-agent system. In 2024, most organizations paid for tokens without measuring them. In 2026, the leading practitioners measure them carefully. By the end of the decade, I expect token governance to be a recognized operational function with its own tooling, its own role definitions, and its own seat at the executive table. The scale of the opportunity warrants the attention. Industry estimates place enterprise spending on large language models at approximately $8.4 billion in 2025, more than double the previous year's figure, with consensus forecasts suggesting another doubling is plausible by the end of 2026. A Deloitte analysis published in early 2026 identifies artificial intelligence as the fastest-growing expense in corporate technology budgets, with some firms reporting that it consumes up to half of their information technology spend. Set against those figures, the waste number bears repeating: 40 to 60 percent of production token budgets is buying nothing at all. The book is a foundation document for the emerging discipline, and it follows the naming precedent that FinOps established: treat tokens as the primary currency of artificial intelligence work and the primary object of measurement and optimization. A Gauge with Five Bands The book organizes the journey as a gauge whose needle sweeps across five bands. Organizations begin in Waste, where consumption is unmeasured. They move to Efficiency as individual habits eliminate the obvious losses, and then to Optimization as routing, caching, and prompt structure become deliberate engineering choices. Leverage follows, when context architecture and agentic design allow the same token to do compounding work across systems. The final band is Advantage, where token discipline becomes an institutional capability that competitors cannot quickly replicate. The subtitle of the book traces the needle's full sweep. The Orchard Principle I frame the practice with what I call the Orchard Principle. Left unattended, a young tree grows enthusiastically in every direction, producing branches, leaves, and complexity, but not necessarily fruit. A master gardener prunes without hesitation, removing healthy branches so the tree directs its energy toward what matters most. Artificial intelligence behaves much the same way. Left unmanaged, prompts become verbose, context accumulates, agents multiply, models are overpowered for simple tasks, and token consumption expands almost invisibly. TokenOps is the discipline of pruning, and the objective is not to consume fewer tokens but to cultivate an ecosystem that produces more insight, better decisions, and greater advantage from every token invested. A wild tree grows; an orchard is designed. What Is Inside The playbook contains sixty-five strategies across seven categories, moving from input and output discipline through format strategy, prompt engineering, model routing, context and caching, and the architecture of agentic systems. The strategies are model-agnostic; they apply whether your teams work in Claude, ChatGPT, Gemini, Grok, or open-source models such as Llama and Qwen, because the underlying mechanics of tokenization, the asymmetric pricing of input and output, and the compounding behavior of context are stable across providers. The book is also built for different readers. An executive can extract the strategic frame and the highest-leverage governance levers in roughly twenty minutes. A technology leader or solution architect will find the structural material on routing, caching, and agentic architecture in the later sections. A hands-on practitioner can begin with ten habits that take fifteen minutes to learn and compound daily thereafter. What the Discipline Returns The evidence base for the effort is substantial. Multiple independent practitioner surveys and production case studies place the realistic savings range for disciplined token usage at 60 to 80 percent of pre-optimization spend, without compromise to output quality. The most disciplined teams, those that layer caching, routing, compression, and retrieval so that the techniques compound, report savings in the range of 80 to 90 percent. The gain is not incremental; it is the difference between AI economics that scale and AI economics that quietly consume the budget innovation needed. Every era of enterprise technology eventually acquires its discipline of stewardship. Cloud found FinOps. Artificial intelligence is finding TokenOps now, and the organizations that build the competence early will hold the same widening advantage that the early FinOps adopters enjoyed. Two hundred and fifty years on, the lesson of 1773 still holds: the remedy for a metered charge you cannot see is not resentment. It is representation, measurement, and a discipline of your own. Optimize every input. Amplify every outcome. And the news gets better. I think most, if not all, of the strategies inside my book TokenOps can be automated. The question of what that looks like for your business or operation is one only you can frame. So I invite you to enjoy your weekend, ponder the significance of Independence Day, and on Monday, dive into this book with the intentional curiosity that leads to action. TokenOps: From Token Waste to Competitive Advantage is available now on Amazon in Kindle, with paperback to follow: https://a.co/d/02NzXwd4 Copyright © 2026 by Severin Sorensen. All rights reserved.

This week the United States turns two hundred and fifty years old. The Declaration of Independence was adopted in the summer of 1776, and so this Fourth of July marks the semiquincentennial, the two hundred and fiftieth year of the American experiment. Two hundred and fifty years is a long time to hold an idea together, and as families gather and fireworks trace the sky, that span deserves more than ceremony. It also deserves a closer look, because this anniversary offers a business writer something he rarely receives, which is two and a half centuries of evidence about what allows an institution to endure. Since my youth I have read and pondered the history of our nation. Ron Chernow’s Washington: A Life taught me about the fragility of the country’s beginning and the character of a man who was great by constitution, by courage, and by his willingness to surrender the reins of power. Ronald C. White’s A. Lincoln: A Biography taught me about the consequence of persistence, about courage under fire, and about the discipline of treating differences among people and ideas as a strength to be harnessed rather than a threat to be suppressed. Doris Kearns Goodwin’s Team of Rivals explored Lincoln’s genius in bringing into his Cabinet men who had opposed him, drawing on their strengths to fashion a more perfect union. Many other biographies have taught me that each of us forms one of the threads that either strengthen or tear the fabric of our society. A democracy must be nurtured, and so must a company if it is to survive its founder. There are limits to any comparison between a country and a company, and I name them plainly here. A company is not a country. Nothing in a strategic plan carries the weight of what it took to bring America to two hundred and fifty years, and a nation holds burdens, a monopoly on force and the involuntary bond of citizenship among them, that no employer ever will. What I am after is therefore a lens rather than an equation. Durable institutions, wherever they appear, tend to face the same structural problems, and the American record throws five of them into unusually clear relief. I offer them not to flatten the nation’s story into a business lesson, for that story is larger and more human than any such lesson can hold, but to draw attention to what mattered before and still matters now. I should also say at the outset what I do not mean by endurance. Survival, by itself, is not a virtue and confers no merit, and a neglected thing can persist for a long while on inertia alone. What we actually mark on an anniversary like this one is not that the union lasted but that, across the years and through trials that might reasonably have ended it, one generation after another chose to tend the framework rather than let it fall. The lasting is the visible residue of that choosing. With that distinction in hand, here are the five patterns. A founding document is not a founding memory The Constitution has been amended twenty-seven times. The document anticipated this, since Article V builds the mechanism for its own revision directly into the framework. The founders wrote an idea meant to be reinterpreted by people who would never meet them, and James Madison, writing as Publius in the Federalist, was candid that they were designing for imperfect people rather than ideal ones, observing that if men were angels no government would be necessary. The framework presumed fallibility and made provision for correction. Most companies do the opposite with their founding story, treating it as a memory to be protected rather than a document to be used. The mission statement is framed and hung in the lobby, recited at the all-hands, and quietly set aside when it becomes inconvenient. The founders of the most durable companies did something different. In Built to Last, Jim Collins and Jerry Porras found that enduring enterprises preserve a fixed core ideology while stimulating change in almost everything else, which is very nearly what Article V does for a republic. Those founders wrote down what the company was actually for, in language specific enough to make real decisions with, and then trusted later leaders to apply it to circumstances they could not have foreseen. The test for anyone inheriting a founding vision is not whether you can recite it but whether you can apply it to a decision the founder never imagined making. Outgrowing the founder is the point The first generation of American leadership was extraordinary, and it was temporary by design. George Washington was urged, in a 1782 letter from Colonel Lewis Nicola, to consider something like a monarchy, and he rebuked the suggestion firmly. He then chose twice to walk away from power, first by resigning his military commission before Congress at Annapolis in 1783, and later by declining a third presidential term. The precedent he set by leaving held for a century and a half on the strength of custom alone, until it was finally written into the Constitution as the Twenty-Second Amendment in 1951. His departures mattered as much as anything he did in office, because they established that the country’s survival could not depend on the continued presence of any single person, however capable. Founders of companies face a quieter version of the same test, and most encounter it without recognizing that they are being tested. The company remains dependent on the founder’s instincts, the founder’s relationships, and the founder’s presence in the room for decisions that ought to have been institutionalized years earlier. This feels like indispensability while it is happening, and it reveals itself as fragility the moment the founder is unavailable, whether by choice, by health, or by ordinary succession. In Good to Great, Collins describes the Level Five leader whose ambition is directed at the institution rather than the self, and who deliberately sets successors up to outperform him. The companies that outlast their founders are led by people who spend their authority building systems, successors, and judgment in others, rather than spending it on remaining necessary. Succession is a discipline, not an event Where the previous pattern concerns the founder’s willingness to step back, this one concerns the machinery that receives what the founder lets go. The United States executed its first genuine transfer of power between rival parties in 1801, when John Adams yielded the presidency to Thomas Jefferson after a bitter contest, and the country has sustained largely peaceful, rule-bound transfers ever since. That record has held across deep disagreement, contested elections, and a civil war in which the nation’s continued existence was genuinely in doubt. It is the product of explicit rules, fixed timelines, and institutions whose legitimacy does not depend on any single outcome being popular, together with what observers call the loser’s consent, the willingness of the defeated to accept the result and contend again another day. Executive succession inside most companies carries none of this rigor. Boards often wait until a chief executive’s departure is imminent or already announced before they treat succession as urgent, and plans exist on paper and nowhere else. The result is that a transition which ought to be the most rehearsed moment in a company’s life frequently becomes its most improvised one. The boards and chief executives who treat succession as a continuous discipline are making the same wager the founders made, that the institution should be stronger than its strongest individual, and that the only way to prove it is to practice the handoff before circumstance forces it. Crisis reveals what was actually built The Civil War stands apart from every other trial in this list, because it alone placed the nation’s survival and its moral foundation in the balance at the same moment. It tested whether the union was a real structure or a convenient arrangement that would dissolve under sufficient pressure, and it did far more than answer that question. It ended chattel slavery, the gravest of the unjust arrangements the young republic had carried, and through the Thirteenth, Fourteenth, and Fifteenth Amendments it redefined national citizenship and the reach of federal authority so completely that the historian Eric Foner has called the period a second founding. It set the nation on the industrial path that would define its next century, from a transcontinental railroad to a unified national currency. All of this was paid for with more than six hundred thousand lives, and by later estimates a good many more. Standing at Gettysburg, Lincoln asked the country to find in that cost a new birth of freedom, and he named the work that remained as unfinished, a description that held long after the guns fell silent. In proving the union real, the war also remade what the union would become. The trials that followed were serious, yet none carried the same peril to the nation’s existence or to its conscience. The Depression tested whether the country’s economic and political institutions could hold when the system meant to provide stability appeared instead to be the source of the instability, and the New Deal and an activist Federal Reserve were among the answers. The financial crisis of 2008 tested the resilience of the banking system and the ingenuity of the Federal Reserve as an independent body willing to act, at times unpopularly, to stabilize the economy. The pandemic tested how far emergency action could reach, and it left a hard lesson about the fiscal consequences that follow when necessary measures in times of crises are not later throttled down after crises. Reasonable people continue to debate the consequences growing national indebtedness, and the widening of what is often called a K-shaped economy in which prosperity and hardship diverge unevenly, and the rise in populism that financial hardship can energize on every side. What each of these episodes has in common is more instructive than their differences. None was survived through the absence of strain. Each was survived through the presence of structures built before the crisis arrived that were strong enough to bend without breaking. Companies discover the same thing during a recession, a leadership scandal, or a sudden loss of market position. A downturn does not create a weak culture; it exposes one that was already weak and had simply never been tested. The trust, the transparency, and the decision-making habits a company relies upon in a crisis cannot be assembled under pressure, because they are the accumulated product of choices made in ordinary times, when nothing appeared to be at stake. The most useful question a leadership team can ask is therefore not how it will respond to the next crisis, but what it has already built, right now, that a crisis would reveal. In the rare crisis grave enough to threaten the enterprise itself, the harder question is not whether you will survive it but whether you have the honesty to be remade by what it exposes. Unity was never the absence of difference The thirteen colonies did not share an economy, a religion, or in many cases much affection for one another. What they shared was a narrow and specific agreement about the handful of things that mattered enough to act on together, paired with wide latitude for everything else. Federalism, whatever its frustrations, was the structural answer to a real problem, which is how genuinely different parties might build something durable together without first being required to become identical. Madison understood that the safeguard lay in structure rather than in virtue alone, writing in Federalist 51 that ambition must be made to counteract ambition, so that the executive, the legislature, and the courts would check one another. Alexis de Tocqueville, observing the young republic in Democracy in America, warned of the tyranny of the majority and concluded that American liberty was sustained less by its laws than by the mores of its people and their habit of forming associations to govern themselves. The best-run companies solve a smaller version of the same problem. A sales organization, an engineering team, and a finance function do not think alike, move at the same speed, or measure success in the same terms, and the attempt to force them into uniformity is usually how a company loses the strength of all three. Leaders who build durable cultures do something closer to what federalism does. They treat the diversity of views as a filter through which ideas are tested against reality and constraint, which strengthens decisions and guards against the quiet conformity of people who follow without the courage to correct. Great companies learn to hold their core values as non-negotiable while inviting creativity, curiosity, and adaptation in the making of the work. The shared principles function like the rails on a winding track, few in number, precise, and permanent, present so that the enterprise can move quickly through everything else without leaving the road. Unity that is built on a small set of shared and well-understood commitments, rather than on enforced sameness, is the kind that tends to hold under load. On endurance, and on what it does not prove It is tempting to treat two hundred and fifty years as proof that the American design was the best of all designs available, that the years themselves have threshed the chaff from the wheat and left only what deserved to remain. I do not believe mere survival can carry that much weight. History offers too many examples of unjust arrangements that endured for generations to let long life stand, by itself, as evidence of virtue. What deserves our respect on this anniversary is not the bare fact that the union lasted but the labor and the sacrifice, real and not imagined, by which it was made and remade. The founding was not a single act but a labor that unfolded across years, precious years. Independence was declared in 1776, yet the framework to sustain it came only later and by degrees, through the Constitutional Convention of 1787, the hard contest of ratification, and the Bill of Rights secured in 1791, with further amendments still to follow. The men who did this work were not united at the outset, and they reasoned their way toward the best structure they could fashion together. Many among them believed they were about more than the construction of an earthly institution, and appeals to Providence run through the founding record, though they were not uniform in their beliefs and their convictions ranged widely. What they held in common was the wisdom to leave within the structure the means of its own correction, so that later generations, if the will of the people so moved them, could amend what the founders had not foreseen. That early framework was fragile, and it remains so today. We build now in stone and steel, and the permanence of our buildings can lull us into forgetting that the government those buildings house rests on nothing so solid. It rests on the continued willingness of many people to hold it together, and at the extremes, in any direction, the structure breaks down. John Adams put the matter starkly in 1798, writing to the Massachusetts militia that our Constitution was made only for a moral and religious people and is wholly inadequate to the government of any other. The machinery presumes a certain kind of citizen and falters without one. A company is no different in this respect. Its charter and its tenets are chosen, and not every enterprise chooses to be good or even lawful. My aim as an executive coach and as a business thinker is to encourage those I work with to seek to do good, to be good, and to lift others, and in their own way help to make a more perfect union of the institutions they lead. The discipline behind the fireworks What we celebrate on this day is not a finished perfection but a tapestry of imperfection, and the honest way to honor it is to take up its improvement on the very day we mark it. America’s two hundred and fifty years belong to a story far larger and more human than any business lesson can hold. Durability, wherever it appears, is a discipline rather than an accident, and this anniversary is as good a moment as any to ask whether we are actually practicing it, in our companies and in our common life alike. Two hundred and fifty is a number, a milestone, a marker set down at one point on a long road. It is not a perfect year, a perfect economy, or a perfect union, and we carry real issues, real disagreements, and real disparities among us. Lincoln, in his first inaugural, appealed to the better angels of our nature, and in his second, with the war not yet ended, he asked the country to act with malice toward none and charity for all, and to bind up the nation’s wounds. That is the spirit I would summon on this day. I trust that if we each answer it, doing our part as individual threads in the tapestry of the nation’s history, we can find the road to compromise and resilience that will carry the country another two hundred and fifty years. So on this anniversary I ask something specific of you and of myself. Lean in, pitch in, and lift the union rather than leave it to chance. Take up the work of a stone builder rather than the posture of a stone thrower. Here is to making our nation’s two hundred and fiftieth a milestone worth remembering, for you and for me, and our children, and our children’s children. Here’s to making it count. Copyright © 2026 by Severin Sorensen. All rights reserved.

For most executives, artificial intelligence has already proven its worth as an analyst, a drafter, and a sounding board. Fewer have discovered what happens when the same conversational fluency is pointed toward something far less utilitarian: music. The same discipline that produces a sharp board memo or a clear strategic brief also produces something unexpected when redirected toward melody, instrumentation, and lyric. Leaders who engage seriously with AI music creation tend to arrive at a similar observation: the tool responds most generously to clarity of intention. It is, in that sense, the newest instrument in a lineage stretching from the first drum to the modern studio. Those who approach it as a creative collaborator find that it extends what they are able to express, without requiring them to become musicians to do so. For business leaders and executive coaches, the most resonant use case is tribute. A song composed in remembrance of a colleague lost, a veteran honored, or a team marking a difficult anniversary can carry meaning that a card or a speech often cannot reach. AI-assisted music creation gives more leaders the practical means to create that kind of work. Music as a reflection of the moment Around the great turning points of the year, many leaders feel the pull to mark the occasion with something beyond routine acknowledgment. Reading history, writing a personal note to the team, gathering people in a room to name what the date represents: these are all worthy practices. In recent years, composing an original piece of music has become a practical addition to that list. With tools like Suno and a disciplined prompting practice, it is now possible to produce a genuine musical reflection of a holiday, one grounded in the specific history being honored and shaped by the emotional register the moment calls for. The process only takes a few hours, requires no musical training, and produces something the audience will remember in a way that slides and speeches typically do not. As the nation marks the 250th anniversary of its independence this week, I’m reminded of songs that I composed for prior American holidays, each one a reflection of a specific moment of remembrance or celebration. They are gathered here because the 250th anniversary of American independence is an appropriate occasion to revisit them. Created at the intersection of AI and human intention, with careful attention to lyric, structure, and emotional register, they honor the men and women in uniform whose service has made this country's history possible. Listen before reading further, and consider what you might create: He Learned to Listen to Understand That Others Live – A Soldier's Last Prayer On That Field of Bomb-Blown Sod – A Soldier's Last Prayer A Soldier's Last Prayer How to create your own song: A step-by-step guide The process of creating a song is more accessible than most leaders expect. It requires no studio equipment and no technical background beyond the prompting skills you already use at work. Here is how it works from start to finish. Step 1: Open Claude or ChatGPT and build your lyrical prompt Begin in conversation with a reasoning model such as Claude or ChatGPT. Describe what you want to create: who the song honors, what quality or memory should anchor it, and the emotional tone you are reaching for. Solemn, resolved, quietly hopeful, and triumphant are all distinct registers that produce meaningfully different outputs. Specify a structure, how many verses, whether a bridge is needed, and what the final chorus should feel like. Include a target length, typically two and a half to three and a half minutes for a piece, and name a genre or reference artist if one comes to mind. Ask the model to produce two outputs: a style description covering instrumentation, tempo, and vocal delivery, as well as a complete lyric sheet with bracketed section labels such as [Intro], [Verse 1], [Chorus], [Verse 2], [Bridge], and [Outro]. Those two elements are what is needed to generate music with precision. Step 2: Copy the lyrics and style description into Suno Navigate to suno.com, where you will need to create a free account to get started. Suno offers a free tier that allows you to generate a limited number of songs, which is enough to explore the tool and produce an initial draft. Once inside, create a new song in Custom Mode, which allows you to supply your own lyrics and style prompt rather than relying on Suno's automatic generation. Paste the style description into the style field, for example: slow acoustic ballad, piano and strings, warm baritone vocal, restrained and building. Paste the lyric sheet, section labels included, into the lyrics field. Suno will typically return two versions simultaneously, giving you an immediate basis for comparison. Step 3: Return to Claude or ChatGPT to fine-tune the lyrics If the lyrics need revision after hearing the initial generation, return to your reasoning model rather than editing inside Suno. Paste in the current lyric and describe what is not working: a chorus that feels too abstract, a bridge that loses momentum, a hook that does not land. Ask for a targeted revision, then bring the updated lyric back into Suno's Remix flow and regenerate. Working iteratively across both tools is where the piece acquires the specificity that generic prompts cannot produce. Step 4: Use "Remix" to refine the version you want When you find a version worth developing further, use Suno's Remix feature to refine it. Remix allows you to adjust specific elements, including instrumentation, tempo, vocal style, and track length, without discarding the version you started with. Two or three passes are usually sufficient to produce something that holds the room. Describe what you want changed in plain language; Suno responds well to specific direction. As always, working with Claude or ChatGPT to refine your Remix prompt before re-entering it into Suno tends to produce better results. Using one AI to prompt another is a reliable pattern across creative workflows, and music generation is no exception. Step 5: Download or Publish the Final Version Once the piece is ready, Suno allows you to download the audio file directly or publish it to the platform. For a tribute intended for a specific event, download the file and play it at the ceremony, embed it in a presentation, or share it with others in advance. For work intended for a wider audience, Suno's distribution options make it possible to publish to Apple Music and other streaming platforms. What makes a strong song The quality of the prompt built in Steps 1 and 4 determines the quality of the output. A few specific choices make a measurable difference. Identify the hook before writing the verses: The hook is the line, typically placed inside the chorus, that the listener carries with them after the song ends. In lyrical work, the most durable hooks name the quality being honored with specificity. For example, courage, steadiness, sacrifice, or service. Starting with that line and building the rest of the lyric outward from it produces more cohesive results than writing linearly from verse to chorus. Genre and instrumentation tags shape the emotional register as directly as the lyrics themselves: A remembrance piece generally calls for restrained acoustic instrumentation at a slow to moderate tempo, with vocal delivery described explicitly in the style field. Specificity in the style description produces noticeably better results than general terms like sad or emotional. Include guidance about space and silence: Instrumental pauses, a sustained note before the final chorus, and a quiet outro rather than a hard stop are choices that give a generated track compositional intention. They can be requested in the prompt and make a significant difference in how the finished piece lands in a room. Working iteratively across multiple tools is where the piece acquires the specificity that generic prompts cannot produce. Leadership lesson beyond the music The value of this practice for executives extends beyond the song. Producing a piece of music through AI requires translating an internal, often inarticulate feeling into language precise enough for a platform to act on. That is a skill with direct applications across every domain where leaders communicate: strategy, feedback, organizational change, and team alignment, and it becomes most apparent in moments where the feeling involved is difficult to name precisely. The leader who can describe grief or gratitude clearly enough to produce a coherent lyric has practiced the same capability required to describe a strategic ambiguity clearly enough for a team to execute against it. Music has always developed alongside the tools available to those who made it. AI-assisted composition is the current chapter in that history. Leaders who engage with it seriously find that the platform rewards the same qualities that good leadership has always required: clarity of intention, precision of language, and a genuine understanding of what the moment calls for. Copyright © 2026 by Severin Sorensen. All rights reserved.

Most executives now accept that artificial intelligence will reshape how their organizations operate. Far fewer have a structured way to build the judgment required to use it well. In March 2026, Anthropic addressed that gap directly by launching Anthropic Academy, a free learning platform built by the company that develops Claude. The Academy is a working curriculum, hosted on Skilljar and linked from anthropic.com/learn, that spans twenty courses across three tracks: AI Fluency for the non-technical leader, Product Training for everyday Claude users, and Developer Deep-Dives for engineers building on the Claude API. Each course is self-paced, free of charge, and ends with a certificate bearing Anthropic's name rather than a third party's interpretation of it. For leaders trying to separate durable AI skill from passing hype, that distinction matters. A curriculum built by the people who build the model tends to age better than one built by an outside observer guessing at how the technology works. Below are six entry points worth knowing, organized roughly from foundational to technical. 1. AI Fluency: Framework and Foundations This is the right starting point for any executive, and it is the course the rest of the Academy builds on. It introduces the 4D framework, Delegation, Description, Discernment, and Diligence, as a discipline for collaborating with AI systems rather than simply prompting them. The course treats fluency as a form of judgment rather than a technical skill, which is precisely the framing a senior leader needs before delegating any consequential work to a model. View the course. 2. AI Fluency for Builders Adapted from the foundational course for people who own outcomes rather than tasks, this version of AI Fluency is built around what Anthropic calls owning the “full arc from problem to shipped solution.” It is suited for operators, founders, and consultants whose value lies in judgment under ambiguity rather than execution of a defined process. View the course. 3. Claude 101 Claude 101 covers the core features most knowledge workers actually need: projects, artifacts, and the practical mechanics of using Claude for everyday work. It is short by design and intended as a baseline rather than a deep dive, which makes it a sensible course to assign broadly across a team before layering in role-specific training. View the course. 4. Introduction to Claude Cowork Cowork is Anthropic's desktop tool for non-developers who want to automate file and task management without writing code, and this course is the clearest introduction to it. It walks through the Cowork task loop, plugins and skills, and how to steer multi-step work responsibly, which is the part most new users underestimate. The course is built for productivity within the first week of use rather than theoretical mastery. View the course. 5. AI Capabilities and Limitations This short, introductory course explains how AI systems actually work, in plain terms, and where their limitations genuinely lie. For executives who need to evaluate vendor claims, set realistic expectations with their boards, or simply avoid embarrassing themselves in a strategy meeting, this is a high-value, low-time-commitment course. View the course. 6. Building with the Claude API For technical teams, this is the flagship deep dive: a comprehensive course covering the full spectrum of working with Anthropic's models through the Claude API, from authentication and request handling to tool use and agentic workflows. It is the course referenced most often by developers moving from chat-based Claude use to building production applications, and it pairs naturally with the Academy's separate courses on the Model Context Protocol for teams connecting Claude to internal systems. View the course. The Main Takeaway None of these six courses requires a paid Claude subscription, and none requires more than an email address to enroll. That is a deliberate choice on Anthropic's part. A better-educated base of users and developers builds more durable applications on Claude, which benefits Anthropic commercially, and a workforce that understands both the capability and the limits of these systems is less likely to misuse them. For leaders building AI fluency inside their own organizations, that alignment of incentives is worth taking advantage of. The full catalog, organized by track, is available at anthropic.skilljar.com, with the discovery hub at anthropic.com/learn. Copyright © 2026 by Severin Sorensen. All rights reserved.

Most chief executives arrive at an executive peer advisory meeting carrying the heavy cognitive residue of everything that preceded it. There is the traffic on the drive over, the urgent unanswered message, the critical board decision left half-made, and the operational firefighting meeting that ran long. The mind is entirely full before the true collaborative conversation has even begun. A thoughtful opening ritual gives members implicit permission to set that fullness down, ensuring the room begins in a state of mindfulness rather than mind-full-ness. While the wordplay is gentle, the performance distinction it marks is operationally significant over time. A robust body of behavioral science and psychological research suggests that the few minutes spent settling attention at the start of a session are among the highest-leverage investments a group can make in its collective intelligence. Here are ten evidence-based opening exercises and affirmations that an Executive Coach or Peer Group Facilitator can draw upon to ground senior leaders. They are designed to be spoken aloud, adapted freely, and rotated over time to keep the practice dynamic rather than dogmatic. Why an Opening Practice Earns Its Place: The Empirical Evidence Before introducing the practical toolkit, we must address the natural skepticism of a results-oriented executive. What is the actual return on a quiet minute? The empirical evidence demonstrates that intentional attention-gathering yields small-to-moderate, yet practically meaningful dividends that compound over time. 1. The Cost of Attention Residue When a leader transitions rapidly between high-stakes contexts, their attention remains divided. This phenomenon, known in organizational psychology as attention residue (Leroy, 2009), dictates that a person’s cognitive capacity is significantly diminished when they switch tasks without a clean mental break—particularly when the prior task involves unfinished or interrupted work. Furthermore, mind-wandering is our baseline default. Killingsworth and Gilbert (2010), utilizing real-time experience-sampling across thousands of participants, discovered that individuals spend roughly 47% of their waking hours thinking about something other than their current activity. Crucially, their data revealed that this chronic mind-wandering actively precedes decreased psychological well-being. For a peer group to unlock breakthrough insights, capturing and anchoring this fragmented attention is a structural prerequisite. 2. Measurable Cognitive Dividends Mindfulness practices directly support executive functioning. In a comprehensive meta-analysis, Zainal and Newman (2024) aggregated 111 randomized controlled trials (RCTs) involving over 9,000 participants. They confirmed that mindfulness-based interventions produce stable improvements in: Global Cognition: Overall clarity and analytical processing. Executive Attention: The ability to intentionally direct focus amidst noise. Working Memory: The mental workspace required to process complex, multi-variable business challenges. Notably, the study found these positive cognitive effects were larger for face-to-face, group-based practices than for isolated, self-guided apps—making a compelling case for embedding these exercises directly into the peer group environment. 3. Neural Systems Regulation & Leadership Agility Under intense pressure, the executive brain can succumb to heightened emotional reactivity, often referred to as "amygdala hijack," where the prefrontal cortex—the seat of logic and long-term planning—is effectively sidelined. Neuroimaging studies demonstrate that mindfulness training alters resting-state functional connectivity, strengthening the coupling between the prefrontal cortex and regulatory pathways (Creswell et al., 2016), which helps down-regulate stress response systems. This aligns cleanly with organizational field data from Rupprecht et al. (2019). In their study of senior leaders completing a workplace mindfulness curriculum, participants reported profound gains in self-reflection, interpersonal attunement, and adaptive change leadership, even when objective cognitive laboratory measures moved modestly. The Pragmatic Takeaway: The scholarly literature does not promise instantaneous transformation from a single breath. It proves that attention is a trainable muscle, presence is a recoverable resource, and a brief opening pause is a hard-nosed investment in a group's collaborative capacity. The Ten Opening Exercises To optimize execution, each exercise below is structured with its psychological mechanism and an explicit facilitator script. 1.1. The Threshold Pause Focus: Combatting Attention Residue. The Mechanism: Leverages spatial psychology to create a clean mental boundary, helping leaders drop the lingering "attention residue" (Leroy, 2009) of prior tasks. Facilitator Script: "As we cross the threshold into this room, let’s take three unhurried breaths together. Whatever complex issues you were dealing with ten minutes ago, visualize leaving them in a box just outside that door. They will be safely waiting for you when we finish. Right now, your value to this room requires your full presence here." 2.2. Interior Weather Checking Focus: Affect Labeling. The Mechanism: Shifting feelings into words—a process called "affect labeling"—has been shown to down-regulate amygdala activity, dampening immediate emotional reactivity (Lieberman et al., 2007). Facilitator Script: "Let’s do a quick round. Share exactly one word that describes your interior psychological weather right now. No explanations, no storytelling, and no attempts from the room to fix it. Just name the state so we can acknowledge it and let it pass." 3.3. The Empty Cup Focus: Reducing Confirmation Bias. The Mechanism: Actively cultivates a "beginner's mind," which combats the cognitive rigidity and confirmation bias that frequently limit highly experienced leaders. Facilitator Script: "There is an old teaching that a cup that is already full can receive nothing new. Most of us are paid to have the answers. For the next few hours, I invite you to intentionally empty your cup. Set aside your certainties so that you have the internal space to be genuinely surprised by the collective wisdom in this room." 4.4. Somatic Grounding Focus: Nervous System Regulation. The Mechanism: Shuts down intellectual overstimulation by temporarily shifting cognitive energy away from circular, stressful thoughts and redirecting it toward physical, sensory data inputs. Facilitator Script: "Leaders live almost entirely in their heads. For the next 45 seconds, let’s return to the body. Feel the absolute weight of your body in your chair. Notice the firm pressure of your feet flat against the floor. Take a deep breath into your stomach, feel the expansion, and let it out slowly. Let your nervous system catch up to your calendar." 5.5. The Strategic Gratitude Round Focus: Optimizing Cognitive Scope. The Mechanism: Reorients attention away from structural deficit thinking. Emmons and McCullough (2003) demonstrated that deliberate gratitude exercises reliably boost positive affect, which broadens a leader's cognitive visual field. Facilitator Script: "To prime our minds for possibility rather than survival-mode problems, let’s do a lightning round. Name one distinct thing, however small or personal, that you are quietly grateful for this week. One sentence maximum." 6.6. The Commitment to Generous Listening Focus: Psychological Safety. The Mechanism: High-quality, non-judgmental listening has been empirically proven to significantly decrease a speaker's defensive processing, leading to greater attitude clarity and superior strategic choices (Itzchakov et al., 2018). Facilitator Script: "The single greatest gift you will offer someone today is your undivided attention. Let’s commit right now to listening entirely to understand, rather than listening to formulate our own clever reply. Let's make this room a rare sanctuary of high-quality attention." 7.7. Exposing the Sub-Problem Focus: Advanced Metacognition. The Mechanism: Encourages metacognition—the capacity to monitor and analyze one's own thinking patterns—which Rupprecht et al. (2019) identified as a premier outcome of mindful leader development. Facilitator Script: "Think briefly about the primary business challenge you brought to the table today. Now, look beneath it. Ask yourself silently: 'What is the real question sitting underneath this presentation?' Hold that deeper inquiry lightly as we begin." 8.8. Radically Lowering the Cost of Candor Focus: Overcoming Impression Management. The Mechanism: Mitigates the cognitive friction of "impression management" (the defensive need to always look competent), allowing genuine, vulnerable data to surface immediately. Facilitator Script: "Out there, you have to be the definitive answer-person for hundreds of employees. In here, you are entirely safe to say: 'I don't know.' Wisdom begins when we drop the armor. Let's make this a room where admitting uncertainty is seen as an act of strategic strength, not weakness." 9.9. The Sixty-Second Still Point Focus: Attentional Recalibration. The Mechanism: Provides a stark, sensory contrast to a day dominated by fragmented interruptions, allowing the brain's attention networks a brief window to consolidate information and reset focus. Facilitator Script: "Before we look at a single agenda item or read a single financial dashboard, we are going to sit together in absolute, deliberate silence for exactly sixty seconds. This minute isn’t empty space; it is the collective mind of this room gathering its power. Let's begin." 10.10. Implementation Intentions Focus: Behavioral Alignment. The Mechanism: Formulating an explicit operational intention maps closely onto Gollwitzer and Sheeran’s (2006) framework, which shows that planning exactly how you will behave produces a medium-to-large effect on actual goal achievement. Facilitator Script: "Close your eyes for five seconds and set a firm, singular intention for how you will show up in this room today. Are you here to be a rigorous challenger? A deeply empathetic listener? Or a leader willing to be uncomfortable and moved? Choose your posture now." The Strategic Why To balance these approaches effectively, a facilitator should map them against the specific emotional energy of the room upon arrival. Remember that while a single session offers an immediate reset, the true cognitive dividends build compound interest through consistent, routine deployment across meetings. Drives psychological safety; builds trust and lowers defensive processing.The group dynamics feel guarded, distant, or overly superficial. A Closing Word for the Chair A peer group facilitator who implements these openings is not asking hard-charging executives to meditate instead of making tough choices. Rather, the facilitator is methodically clearing the cognitive pipeline through which excellent choices travel. Hold these practices with the strategic humility they demand. The individual effects of a brief mindfulness practice are modest, but they are incremental and aggregate beautifully over time. Start with mindfulness, and the "mind-full-ness" that occurs later in your agenda will be the precise, high-yielding, creative kind that drives enterprise value. References Creswell, J. D., Taren, A. A., Wager, T. D., et al. (2016). Alterations in resting-state functional connectivity link mindfulness meditation with reduced biomarkers of inflammation: A randomized controlled trial. Biological Psychiatry, 80(1), 53–61. https://doi.org/10.1016/j.biopsych.2016.01.017 Emmons, R. A., & McCullough, M. E. (2003). Counting blessings versus burdens: An experimental investigation of gratitude and subjective well-being in daily life. Journal of Personality and Social Psychology, 84(2), 377–389. https://doi.org/10.1037/0022-3514.84.2.377 Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119. https://doi.org/10.1016/S0065-2601(06)38002-1 Itzchakov, G., DeMarree, K. G., Kluger, A. N., & Turjeman-Levi, Y. (2018). The listener sets the tone: High-quality listening increases attitude clarity and behavior-intention consequences. Personality and Social Psychology Bulletin, 44(5), 762–778. https://doi.org/10.1177/0146167217747874 Kabat-Zinn, J. (1994). Wherever you go, there you are: Mindfulness meditation in everyday life. Hyperion. Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 330(6006), 932. https://doi.org/10.1126/science.1192439 Leroy, S. (2009). Why is it so hard to do my work? The challenge of attention residue when switching between work tasks. Organization Science, 20(2), 168–181. https://doi.org/10.1287/orsc.1080.0379 Lieberman, M. D., Eisenberger, N. I., Crockett, M. J., et al. (2007). Putting feelings into words: Affect labeling disrupts amygdala activity in response to affective stimuli. Psychological Science, 18(5), 421–428. https://doi.org/10.1111/j.1467-9280.2007.01916.x Rupprecht, S., Falke, P., Kohls, N., et al. (2019). Mindful leader development: How leaders experience the effects of mindfulness training on leader capabilities. Frontiers in Psychology, 10, 1081. https://doi.org/10.3389/fpsyg.2019.01081 Zainal, N. H., & Newman, M. G. (2024). Mindfulness enhances cognitive functioning: A meta-analysis of 111 randomized controlled trials. Health Psychology Review, 18(2), 369–395. https://doi.org/10.1080/17437199.2023.2248222 Copyright © 2026 by Severin Sorensen. All rights reserved.

Most executives know the pre-mortem. Very few use it. The concept, developed by psychologist Gary Klein and popularized in organizational strategy circles by Daniel Kahneman, is disarmingly simple. Before committing to a major decision, you imagine it is twelve months in the future and the initiative has failed catastrophically. You then work backward to explain what went wrong. The exercise forces a team to surface its private doubts, challenge its shared assumptions, and confront the risks it had been too optimistic to name. The reason executives know the pre-mortem but rarely use it is not a lack of appreciation for its value. It is a lack of time. Running a rigorous pre-mortem requires facilitation, honest conversation, and protected space on a calendar that is already overcrowded. The result is that most leaders move forward with hope as their primary risk management strategy. AI eliminates that excuse entirely. What the Pre-Mortem Was Always Meant to Do Klein's original insight was that human beings are naturally inclined toward optimism when they are invested in a plan. The psychological phenomenon he identified causes teams to underweight the probability of failure and overweight the quality of their own preparation. The pre-mortem was designed to create a structured permission structure for pessimism: a moment in which raising concerns was not only acceptable but expected. For executive coaches, this matters because the leaders they work with are frequently the most optimistic people in any room. They have been selected, promoted, and rewarded for their belief in what is possible. That same quality that makes them effective leaders also makes them systematically vulnerable to overlooking what could go wrong. Coaching that fails to surface that vulnerability leaves the executive exposed. Where AI Changes the Equation An AI system has no emotional investment in the plan you are evaluating. It carries no political allegiance to the executive who championed it, no loyalty to the team that built it, and no career risk from naming the possibility of failure. When prompted thoughtfully, it will generate failure scenarios with a thoroughness and dispassion that no internal team member can easily replicate. "You are a strategic thinking partner with deep experience in organizational risk analysis and executive decision-making. I am a senior leader who is about to commit to a significant initiative, and I want us to work through the risks together before I move forward. Where my description of the situation is incomplete, ask me clarifying questions before drawing conclusions. Your tone should be analytical but constructive, the kind of honest assessment a trusted advisor would offer before a high-stakes commitment. Our purpose is to surface the failure modes I have not yet named, so that I can make a better decision before momentum makes it harder to course-correct. To anchor your thinking, treat this as a situation where the initiative has visible executive sponsorship, is moderately well-resourced, and has already begun building internal support. With that context in mind, let us begin: assume it is eighteen months from now and this initiative has failed significantly enough to affect my organization's credibility with key stakeholders. Before generating any explanations, ask me the three questions that would most sharpen your analysis of what went wrong. Then, once I have answered, provide the ten most plausible failure scenarios in order of likelihood, and for each one identify the early warning signal that should have been visible at the outset. Present your findings in a format I can bring into a conversation with my leadership team. Here’s the initiative and situation: [insert details here]." In under ten minutes, a leader will have a failure analysis that would have taken a two-hour facilitated session to produce with a human team, and that session would still have been filtered through the political dynamics of the room. The AI output is not the final answer. It is the starting point for a sharper, more honest conversation. The coach's role shifts from facilitating the discovery of concerns to helping the leader evaluate which concerns are most material and what commitments they are willing to make in response. The Three Failures AI Catches That Teams Miss In practice, the AI-accelerated pre-mortem tends to surface three categories of risk that internal teams consistently underweigh. Execution risk at the edges of accountability. Most strategic plans assign ownership for the core deliverables and leave the interdependencies between functions to chance. AI consistently identifies the handoff points, the shared assumptions between teams, and the places where everyone assumes someone else is responsible. Market and timing assumptions. Plans built during a period of organizational confidence often embed assumptions about external conditions that are never explicitly stated. AI will name those assumptions and ask what happens if they do not hold. Leadership capacity and bandwidth. Perhaps the most consistently overlooked failure mode is simply that the people responsible for executing the plan are already fully committed elsewhere. AI will identify this pattern with notable regularity because it has no interest in flattering the leader's confidence in their team's capacity. The Standard Has Changed Executives who are not incorporating AI into their initiative preparation are working at a fraction of their potential. The tool does not replace the executive’s judgment, it removes the logistical barrier that has kept the pre-mortem from becoming standard practice for the leaders who need it most. Executives are making consequential decisions every week. Most of those decisions are moving forward without a structured failure analysis. AI makes that analysis available in the time it takes to draft the prompt. The pre-mortem was always a good idea. Now there is no longer a good reason not to use it. References Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. ISBN: 978-0374275631 Klein, G. (2007). "Performing a Project Premortem." Harvard Business Review, 85(9), pp. 18–19. Available at: hbr.org/2007/09/performing-a-project-premortem Copyright © 2026 by Severin Sorensen. All rights reserved.

By the time you bring a decision to a conversation, you have usually already made up your mind. You may not realize it. The situation may feel genuinely open, the options unresolved. But in most cases, the moment you begin describing a problem to another person, you have already climbed a cognitive staircase that began with a selective observation and ended with a firmly held conclusion, moving through layers of interpretation, assumption, and belief so quickly that the process left no visible trace. Chris Argyris named this staircase the Ladder of Inference. Peter Senge brought it to a broader organizational audience through The Fifth Discipline. The model is well known in leadership development circles. The challenge has always been personal: how do you see the rungs you skipped when the climb happened automatically and the view from the top feels like obvious truth? AI provides a practical answer to that question for the first time. A Brief Account of the Ladder Argyris described the Ladder of Inference as the mental pathway through which human beings move from raw data to action. The progression begins at the observable facts of a situation and moves upward through a series of interpretive steps: we select the data that seems relevant, we interpret what that data means based on prior experience, we form assumptions from those interpretations, we draw conclusions from our assumptions, we build or reinforce our beliefs from those conclusions, and finally we take action based on those beliefs. The problem is not the process. The problem is the speed and the invisibility of it. By the time you frame a situation as requiring a decision, you have already selected your data, interpreted it through your existing mental models, and formed a conclusion that now presents itself as self-evident. The early rungs of the ladder are gone from view. Argyris further observed that individuals tend to operate from a reflexive loop in which their beliefs about the world determine which data they select, reinforcing those same beliefs over time. Leaders who are never asked to examine the bottom rungs of their own ladder become progressively more confident and progressively less accurate. The more experience you accumulate, the more invisible this pattern becomes. What AI Can Do That a Trusted Advisor Cannot Easily Do Alone A skilled advisor, peer, or coach can ask powerful questions that slow you down and encourage reflection on your reasoning. What no human interlocutor can do efficiently, in real time, with the full breadth of a situation in view, is independently reconstruct your inferential chain from the available facts and name, in specific terms, where your account diverges from the observable data. This is precisely where AI becomes a powerful thinking instrument. When you describe a situation to an AI system and ask it to separate observable data from interpretation and assumption, the output creates a map of the ladder that you could not have produced as quickly or as dispassionately on your own. The AI has no stake in your conclusion, no relationship to protect, and no career risk from telling you that your reasoning rests on something that has not been verified. Using the WHISPER Framework leveraged in PromptSensei, for structured AI collaboration, an executive might open this dialogue as follows: "You are a strategic thinking partner trained in organizational behavior and decision science, with a specific focus on the Ladder of Inference developed by Chris Argyris. I am a senior executive working through a high-stakes situation, and I want us to examine my reasoning together before I act. Where my account is incomplete or where you need clarification to do this accurately, ask me before drawing any conclusions. Your tone should be analytically honest and direct, the kind of assessment a trusted advisor with no personal stake in the outcome would offer. Our purpose is to separate what I can actually observe from what I am interpreting or assuming, so that I can identify where my reasoning may be carrying more weight than my evidence warrants. I will describe the situation in full. Once I have, please organize my account into three categories: what is directly observable or verifiable, what represents my interpretation of those observations, and what appears to be an assumption I am treating as established fact. Where an interpretation or assumption seems to be doing significant work in my reasoning, flag it and ask me what evidence I am drawing on. Present your analysis in a format I can sit with before my next major decision. Here is the situation: [insert details here]." The output will not tell you what to decide. It will show you the structure of how you arrived at your current position, including the places where your confidence exceeds your evidence. That is the information most executives never see, because most of the people around them are too invested in the outcome to name it. Three Patterns That Surface Repeatedly In working with executives who use AI-assisted ladder analysis, three assumption patterns appear with notable consistency. Attribution without evidence. Executives frequently ascribe intentions, motivations, or attitudes to others based on behavior alone. A board member who asked a pointed question in a meeting is characterized as adversarial. A direct report who missed a deadline is described as disengaged. The behavior is observable. The attribution is an inference, and it is often carrying enormous weight in the decisions that follow. Selective data framing. Under pressure, it is natural to present the evidence that supports the conclusion you have already reached and to omit the evidence that complicates it, often without realizing you are doing so. AI, given the full account, will identify when your reasoning would change materially if certain facts were included. Generalizations from single events. A single failed initiative becomes evidence that the organization cannot execute. A single difficult quarter becomes evidence that the strategy is fundamentally flawed. AI will identify when a conclusion of significant scope is being drawn from a data set too narrow to support it. The Shift This Creates in How You Lead When you examine your own reasoning before acting, rather than after a decision has gone wrong, the dynamic of every subsequent conversation changes. You arrive at the table having already interrogated your assumptions. You know which parts of your position are grounded in evidence and which parts are grounded in belief. That self-awareness sharpens your authority. The questions you ask of your team will be more precise. The moments when you choose to hold your position or revise it will be better calibrated. The probability of reaching the decisions that prove durable over time increases substantially, because you are no longer confusing confidence with clarity. Argyris spent decades arguing that the Ladder of Inference was one of the most important tools available for helping intelligent people learn from experience. The barrier has always been practical: surfacing the ladder requires time, honesty, and someone with no stake in your conclusions. AI provides exactly that, and it does so before the decision is made. References Argyris, C. (1990). Overcoming Organizational Defenses: Facilitating Organizational Learning. Pearson Education / Prentice Hall. ISBN: 978-0205123384 Senge, P. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. Doubleday. ISBN: 978-0385517256 Copyright © 2026 by Severin Sorensen. All rights reserved.