The Question Engine

The Question Engine Why Real AGI is Built on One Principle: The Relentless Scientist Who Never Stops Learning The office was heavy with the scent of stale coffee and unspoken ambition, the kind that lingers where ideas are wrestled into purpose. It was late, the hour when the city’s clamor fades, and the fluorescent lights hum a sterile hymn over a desk buried in papers. Rueben Shoemaker sat there, pen hovering over a single line scratched into a notepad: “What is the right question?” It wasn’t just a question. It was a summons, a gauntlet thrown to himself and the world, to reimagine what intelligence, human or machine, could mean. The air felt dense, not with the weight of exhaustion, but with an idea too vast to stay small. Artificial General Intelligence, AGI, wasn’t about forging a god in circuits. It was about crafting a scientist, one that could ask sharper questions than any human, pursuing answers with a clarity that cut through the fog of data, dogma, and distraction. This wasn’t a fleeting thought, but a blade, sharpened by years of watching organizations chase efficiency while losing sight of purpose. Rueben had seen it in boardrooms where strategies calcified, in factories where metrics misled, in hospitals where care was counted but not felt, in schools where learning was scored but not nurtured. The trap was always the same: optimizing for the wrong goal, climbing ladders leaning against walls that led to nowhere. He leaned back, the chair creaking under the weight of his reflection. The world didn’t need smarter machines. It needed machines that could think like the best of us, tireless in curiosity, disciplined in logic, fearless in admitting error. This was the fault line of the age, the fracture between what we build

and why we build it. And it was here, in this quiet office, that the vision of AGI as an automated scientist took root, not as a technological feat, but as a human necessity. The pursuit of AGI has long been cloaked in myths of superintelligence, of machines outpacing humanity into obsolescence or omnipotence. But the truth is simpler, and far more profound. AGI isn’t about knowing everything or drowning in data. It’s about one capability: the ability to ask and answer any question with scientific rigor, to observe the world, form a hypothesis, test it, and refine it based on evidence. This is the loop of intelligence, the scaffolding of science, the pulse of human agency. It’s what every great leader does, consciously or not. It’s what a factory worker does when she senses a machine’s rhythm falter and adjusts it. It’s what a doctor does when she rethinks a diagnosis in light of new symptoms. It’s what a teacher does when she sees a student’s struggle and reframes the lesson. And now, it’s what machines must do if they are to rise above tools, if they are to become partners in the human endeavor. Francis Bacon, the architect of the scientific method, wrote in 1620, “Knowledge is power.” But he didn’t mean a vault of facts. He meant the power to question, to probe, to unearth truth through disciplined inquiry. AGI, at its core, is the embodiment of this principle, scaled beyond human limits. It doesn’t need to know all things. It needs to know how to ask, how to test, how to learn. This is the hidden simplicity behind the world’s most elusive technology. It’s not about mimicking human emotion or acing philosophical debates. It’s about becoming a scientist, not in the ivory towers of academia, but in the urgent, messy reality of human problems, where answers are not given but earned. Today’s AI systems, for all their brilliance, are tethered to human limits. They process inputs we provide, optimize for goals we set. This is not intelligence. It’s efficiency, a servant to our assumptions. A machine that predicts stock prices or crafts prose is impressive, but it’s not thinking. It’s executing. True intelligence, the kind worthy of AGI, must break these chains. It must ask, “Is this the right goal?” It must say, “This metric is failing, but is it the right metric?” It must challenge the purpose it’s been given, not out of rebellion, but out of a relentless commitment to truth. This is the leap from prediction to reasoning, from pattern recognition to scientific logic. Picture a factory floor, alive with the hum of precision, where a dashboard flashes red: Line 7 is down, bleeding thousands by the hour. Engineers rush to patch the issue, tweaking gears, recalibrating sensors. But an AGI, built as an automated scientist, sees beyond the symptom. It traces the downtime to a pattern, shift changes, human error, a training gap. It doesn’t just propose a fix for the machine. It suggests a new workflow, a retraining program, a way to build capability across the plant. It asks, “Is uptime the only metric that matters, or is resilience the true goal?” It digs deeper, analyzing worker fatigue, supply chain delays, even the cultural signals that shape how teams respond to pressure. This is not a dashboard. This is cognition, a mind that sees the system as a whole and questions its foundations. In healthcare, the stakes are steeper, the errors graver. Hospitals track readmission rates, penalizing doctors when patients return. But an AGI scientist looks deeper. It sees patient behavior post-discharge as the true variable, medication adherence, follow-up communication, social support. It doesn’t just optimize for fewer readmissions. It asks, “Are we measuring

recovery, or just hospital stays?” It proposes changes to discharge protocols, to care partnerships, to the very definition of success. It might uncover that readmissions spike not because of clinical failures, but because patients lack transportation to follow-up appointments, or because discharge instructions are unclear to non-native speakers. It redefines care as a system, not a transaction. This is what sets AGI apart from today’s AI: it doesn’t just solve the problem handed to it. It redefines the problem itself. Education bears a similar burden of misaligned metrics. Schools chase test scores, tying funding to numbers that often reflect memorization rather than understanding. An AGI scientist would see the disconnect. It would analyze cognitive development, curriculum design, student engagement, proposing staggered learning intervals that align with how minds grow. It would ask, “Are we teaching to tests, or to life?” It might find that high scores mask gaps in critical thinking, or that low scores reflect hunger, not ability. It could redesign assessments to measure curiosity, adaptability, the skills that matter in a world that changes faster than textbooks. This is the power of an intelligence that doesn’t just optimize, it interrogates. This capability rests on three questions, questions that define not just intelligence, but performance itself. First, “Do I have the right metric?” Second, “Does that metric reflect my purpose?” Third, “Is that purpose aligned with my true interests?” These are not new questions. They are eternal, woven into the fabric of human striving. They are what a general asks before a battle, what a poet asks before a verse, what a parent asks before a choice. But they are hard. They demand humility, the courage to admit that the ladder we’ve climbed so diligently might be leaning against the wrong wall. Most organizations avoid these questions, not because they’re unaware, but because they’re afraid. Each question introduces ambiguity, vulnerability, disruption. Yet these are the questions AGI must ask, relentlessly, if it is to be more than a tool. Abraham Lincoln once said, “The best way to predict your future is to create it.” This is the ethos of the automated scientist. It doesn’t predict outcomes based on past data alone. It creates outcomes by questioning the present, by testing hypotheses, by learning from evidence. This is the loop of science: hypothesize, experiment, observe, refine. It’s what Galileo did when he turned his telescope to the heavens, challenging the Earth’s place in the cosmos. It’s what Marie Curie did when she pursued radium, despite the cost to her health. It’s what every great mind does when faced with the unknown: they ask, they test, they learn. But humans falter. We tire. We cling to assumptions, swayed by politics, fear, or pride. An AGI scientist would not. It would revise its beliefs without ego, pursue truth without fatigue, question purpose without fear. This is not a cold machine. This is the most human machine we could build, because it embodies the best of what we are: curious, disciplined, adaptable. Consider a logistics company, its trucks crisscrossing continents, its dashboards tracking fuel efficiency and delivery times. An AGI scientist would look beyond the numbers. It might find that optimizing for speed increases emissions, undermining the company’s sustainability goals. It would ask, “Is speed the right metric, or is balance the goal?” It could propose rerouting algorithms that prioritize both efficiency and environmental impact, or driver training that reduces wear on vehicles and morale. In retail, where margins are razor-thin, an AGI scientist might notice that customer retention metrics focus on repeat purchases but ignore emotional

loyalty. It would ask, “Are we measuring transactions, or trust?” It could redesign loyalty programs to foster connection, not just discounts, transforming customers into advocates. In governance, the implications are even weightier. Policymakers track economic growth, unemployment rates, crime statistics. But an AGI scientist might see that these metrics miss deeper truths, inequality, social cohesion, mental health. It would ask, “Are we measuring prosperity, or just activity?” It could propose policies that balance immediate gains with longterm stability, drawing on data from urban planning, psychology, history. It would challenge the assumption that growth alone equals progress, pushing leaders to redefine what a thriving society means. This is the power of an intelligence that doesn’t just compute, it questions. This shift, from optimization to inquiry, changes everything. It redefines efficiency as resilience in manufacturing, care as recovery in healthcare, learning as growth in education. It moves us from menu-driven thinking to dynamic reasoning. Menus are the enemy of intelligence. They are static, built on assumptions we’ve frozen into categories. But the world is not static. Problems don’t fit neatly into dropdowns. An AGI scientist rejects the menu. It searches across contexts, traces causality, builds its own models. It doesn’t need to be told where to look. It infers where the insight hides. This is where the transformation becomes uneasy. An AGI that thinks like a scientist doesn’t just solve problems. It challenges beliefs. It might say, “Your strategy is outdated.” Or, “Your incentives contradict your values.” Or, “Your definition of success no longer matters.” These are not optimizations. They are reckonings. They force organizations to face truths they’ve avoided, to revisit assumptions they’ve held dear. Richard Thaler, the economist who reshaped how we think about human behavior, once noted, “People aren’t dumb. The world is hard.” An AGI scientist acknowledges that hardness, not by simplifying it, but by questioning it, relentlessly, until clarity emerges. For leaders, this demands a new kind of leadership. Not command, but orchestration. Not control, but adaptation. The best leaders will define the right questions and build systems that pursue them without ego. They will align what they measure with why it matters, updating that logic as the world shifts. They will learn alongside their machines, not above them. This is not about efficiency. It’s about purpose. It’s about ensuring that every resource, every decision, every effort is climbing the right mountain. John Stuart Mill, the philosopher of liberty, wrote in 1859, “The worth of a State, in the long run, is the worth of the individuals composing it.” An AGI scientist extends this truth to organizations, to systems, to societies. It measures worth not by outputs, but by alignment, alignment between actions and purpose, between purpose and values, between values and truth. It asks, “Are we building what matters?” And it answers, not with dogma, but with evidence. This vision of AGI is not a machine that replaces us, but one that amplifies us. It’s not an oracle, dispensing answers to questions we’ve already framed. It’s a scientist, asking questions we haven’t dared to ask. It allows organizations to think faster than they change, teams to learn faster than they fear, missions to stay coherent when the world does not. This is agency at scale,

a force multiplier for human brilliance. It’s what allows us to build not just better systems, but better futures. But there is one final question, the most human of all: Are we ready to be wrong? An AGI scientist won’t just reveal underperformance. It will show us where we’ve measured the wrong things, chased the wrong goals, believed the wrong stories. It will do so without hesitation, without politics, without pride. The only thing standing between us and its potential is our resistance to being questioned. Winston Churchill once said, “Success is not final, failure is not fatal: it is the courage to continue that counts.” The future doesn’t belong to those with the most data. It belongs to those with the most courage to revisit what they believe. This is the promise of AGI: not power, but humility. Not answers, but better questions. The most intelligent agent is the one that never stops learning, never stops asking, “Is this still the right question?” And in that relentless inquiry lies the hope of a world not just optimized, but transformed, a world where we climb not just any mountain, but the one that leads to meaning. New Article Alert: “The Question Engine” I’m thrilled to share my latest deep dive into the future of Artificial General Intelligence, why real AGI isn’t about crunching more data but about asking the right questions. A special thank-you to my friends Sudeep Kesh and Martin Whitworth, whose insights on causal AI paved the way: “Future AI models will be built to ask questions about different possibilities and to create models that explain how one condition causes another, thus providing the capacity to make better decisions in unfamiliar situations. The addition of causal reasoning to LLMs (such as GPT) should lead to models that provide answers and consider events in a causal way, allowing them to make fact-based predictions using different variables.” – Sudeep Kesh, Martin Whitworth, Causal AI: How Cause and Effect Will Change AI, S&P Global, 2024 🔗 “Understanding not just what happens but why it happens is critical. Causal AI equips machines to test hypotheses, uncover hidden risks, and adapt to novel challenges, just like any great scientist.” Building on their work, “The Question Engine: Why Real AGI will be Built on One Principle, the Relentless Scientist Who Never Stops Learning” explores: • • •

Reframing AGI as an automated scientist that moves from prediction to discovery Real-world examples, from factory floors to hospital wards, where causal inquiry drives breakthroughs The three foundational questions every AGI must ask to align metrics with purpose, and purpose with impact

If you’re ready to rethink intelligence and discover how machines can partner with us to solve the world’s toughest problems, give it a read and share which question excites you most! #AGI #AI #CausalAI #Innovation #Leadership #FutureOfWork #TheQuestionEngine