“Logic will get you from A to B. Imagination will take you everywhere.”
The line is often attributed to Einstein. Whether or not the attribution is fully settled, the idea feels newly relevant in the age of AI. Logic has always helped us move from one step to the next. Imagination helped us question the destination, reframe the problem and see possibilities beyond the obvious path.

AI does not weaken the point. It intensifies it.
Logic with AI will get you from A to B faster. Imagination with AI will take you everywhere.

That is the promise. And the risk.

AI can help us move faster through analysis, synthesis, comparison, writing, simulation and decision preparation. It can produce the report, prepare the meeting, scan the market, rewrite the email, generate the options and create the deck. But it does not automatically tell us whether B is the right destination. It does not automatically know whether the problem has been framed well, whether the assumptions are sound, whether the path is worth taking, or whether the answer should change the decision.

This is the leadership problem hidden behind the current enthusiasm for AI. Most organizations are still treating AI as a better answer machine. That is understandable. The productivity gain is visible. People can write faster, research faster, prepare faster and produce faster. But speed is not the same as intelligence. A fast answer to the wrong question can create a very efficient form of strategic waste.

AI can make weak thinking look polished. It can turn vague assumptions into fluent analysis. It can create professional-looking output from poorly framed problems. It can accelerate consensus around a question that should have been challenged first.

That is why the next AI advantage will not belong to organizations that simply ask AI for more answers. It will belong to leaders and teams that ask better questions before everyone else.

For most of modern business, answers were expensive. Research took time. Analysis required expertise. Reports required teams. Benchmarking required consultants. Market scans required agencies. Strategy decks required weeks of synthesis. Even getting a clear internal view of customers, competitors, operations or performance could require long cycles of data collection, interviews and coordination.

AI changes that economics. Answers become cheaper, faster and easier to generate. A manager can ask for a market analysis. A marketer can request five campaign angles. A sales team can generate an account brief. A product leader can summarize customer feedback. A CEO can ask for strategic scenarios. A consultant can create a first version of a board memo in minutes.

This is useful. But it also creates a new problem. When answers become abundant, they stop being the main source of advantage. The scarce capability moves upstream. It becomes the ability to define the question, frame the problem, challenge the premise, select the right evidence, interpret uncertainty and decide what the answer is supposed to change.

In other words, AI reduces the cost of producing answers. It increases the value of thinking well.

AI does not eliminate the need for strategic thinking. It raises the standard. As answers become faster and more abundant, competitive advantage shifts toward question quality: the ability to frame the right problem, expose hidden assumptions, guide AI with context, test alternative explanations and connect insight to decisions. Weak questions lead to fast but shallow answers. Better questions create better executive conversations, sharper choices and stronger execution. The future leadership capability is not prompt engineering alone. It is problem framing, judgment and disciplined inquiry.

Many organizations still behave as if their main problem is information access. They ask for more data, more dashboards, more reports, more market intelligence, more customer insight and now more AI-generated synthesis. But in many leadership teams, the real bottleneck is not information. It is interpretation.

The organization may already have the facts. Sales knows where customers hesitate. Marketing sees which messages underperform. Customer service hears recurring frustrations. Product knows which features are underused. Finance sees margin pressure. Retail teams know where the shelf story breaks. Employees know which workflows create friction. The problem is that these signals are scattered, softened, delayed or interpreted through existing assumptions.

AI can help connect those signals, but only if leaders ask the right questions.

“Summarize customer feedback” is useful. But “which customer objections are increasing, which of them threaten our value proposition, and what should we change in sales enablement next month?” is better.

“Analyze competitors” is useful. But “which competitor claims are reframing the category in a way that could make our portfolio look less relevant?” is better.

“Create a growth strategy” is too vague. But “where are we leaking growth between demand generation, conversion, retention and portfolio mix?” is much sharper.

The quality of AI output depends heavily on the quality of the frame. Poor framing produces generic answers. Strong framing produces decision-relevant intelligence.

AI can make bad thinking faster in three ways.

First, it can amplify hidden assumptions. If a team asks, “How can we launch this product faster?” AI may help create a launch plan. But the better question may be whether the product is ready to launch, whether the customer need is strong enough, whether the value proposition is clear, whether the sales team can explain it, or whether the launch should be narrowed to a specific segment first. A weak question assumes the direction. A strong question tests it.

Second, AI can create false completeness. A well-structured answer can feel more reliable than it is. It may include frameworks, bullet points, scenarios and confident recommendations. That polish can reduce healthy skepticism. Leaders may mistake fluent synthesis for deep understanding. The issue is not that AI is useless. The issue is that the human reader must remain alert to what is missing, uncertain, biased, outdated or untested.

Third, AI can accelerate organizational theatre. Teams can produce more plans, more documents, more dashboards and more analysis without changing decisions. The organization looks smarter, but the work does not move. Reports become sharper. Meetings become better prepared. Presentations become more polished. Yet the same slow choices, unclear priorities and weak execution loops remain.

AI is powerful. That is precisely why weak thinking becomes more expensive.

There is a risk that companies reduce this topic to prompt engineering. Better prompts help. People need to learn how to give context, define outputs, set constraints, ask for alternatives, request critique, test assumptions and iterate. But prompt engineering is only the surface layer.

Question quality is deeper. It is the ability to understand what problem deserves attention, what decision is at stake, what evidence is relevant, what assumptions are hidden, what trade-offs matter, what risks are being ignored and what action should follow.

A good prompt can improve an AI answer. A good question can improve the business conversation.

For leaders, this distinction matters. The future is not only about employees becoming better AI users. It is about organizations becoming better thinkers. That requires routines where questions are treated as strategic assets: in executive meetings, strategy reviews, portfolio discussions, go-to-market planning, AI pilots, customer reviews and performance conversations.

A company with poor questions will use AI to accelerate noise. A company with sharper questions will use AI to expose reality earlier.

AI will make reports easier to produce. That may be useful, but it also reduces the value of reports as a differentiator. If every organization can generate a market scan, competitor overview, customer summary or strategic memo, the advantage will not be the report itself. It will be the conversation the report enables.

This is an important shift. Many organizations still treat information as the deliverable. The deck is completed. The dashboard is updated. The report is circulated. The meeting is held. But the real value comes only when information changes the quality of decision-making. What tension does it reveal? What assumption does it challenge? What choice does it force? What priority does it sharpen? What action does it trigger?

AI can help leaders prepare better conversations. It can surface contradictions, compare scenarios, generate counterarguments, summarize weak signals and identify decision options. But leaders must design the conversation around the right questions. Otherwise, AI only creates better-looking material for the same slow decision system.

This is where the future consultant, advisor or strategy leader changes. The value shifts from being the person with the answer to being the person who can architect the conversation that produces a better decision.

Before asking AI for an answer, leaders should ask five human questions.

The first is: what decision will this answer inform? If there is no decision, the output risks becoming intellectual activity. AI can generate impressive analysis, but business value appears only when analysis changes a choice, an action or a behavior.

The second is: what assumption are we making? Every question carries assumptions. Are we assuming the market wants the product? Are we assuming growth is blocked by demand rather than conversion? Are we assuming AI should automate the task rather than redesign the workflow? Are we assuming the customer sees the category as we do?

The third is: what would change our mind? This question protects leaders from confirmation bias. AI can easily be used to support a preferred view. Better leaders use it to challenge the view, identify disconfirming evidence and explore alternative explanations.

The fourth is: which context does AI need to answer well? Generic questions produce generic outputs. Strong questions include strategic context, customer segments, constraints, current performance, competitive dynamics, prior decisions and desired trade-offs.

The fifth is: what action should follow? AI outputs should not end as documents. They should feed decisions, experiments, changes in messaging, workflow redesign, customer action, portfolio choices or execution priorities.

These five questions do not slow AI down. They prevent speed from becoming waste.

The Einstein-attributed line is useful because it reminds us that problem solving is not only about reaching a known destination efficiently. Logic helps us move from one step to the next. AI can accelerate that logic dramatically. But imagination helps us ask whether the path is the right one, whether the destination should change, whether the category is being framed too narrowly, whether the customer problem is different from what we assumed, or whether the organization is solving yesterday’s issue faster.

In the AI era, imagination is not decoration. It is strategic capacity. It is the ability to reframe. For example, a consumer electronics brand might ask, “How do we communicate our AI features better?” A more imaginative question is, “Which customer problem does our AI actually make easier, and is that problem strong enough to justify a premium?”

A commercial team might ask, “How do we generate more leads?” A stronger question is, “Where does demand already exist, but fail to convert because the value proposition, channel execution or sales narrative is weak?”

A CEO might ask, “How do we make our teams more productive with AI?” A better question is, “Which workflows should no longer exist in their current form now that AI is available?”

AI can help answer each of these. But only the better question opens the better strategic space.

The title of this article matters because it captures the real shift.

AI gets you from A to B faster. That is productivity. It is useful. It is measurable. It is often the first benefit organizations see.

But better questions decide where to go. That is strategy. It is harder to measure, but far more consequential.

A company can use AI to accelerate a weak plan. It can use AI to produce more polished arguments for yesterday’s assumptions. It can use AI to make internal reporting faster without improving decisions. It can use AI to automate tasks that should have been redesigned. It can use AI to scale messaging that customers no longer believe.

Or it can use AI to challenge the destination.

Why are we going to B? Is B still relevant? What has changed in the market? What customer problem are we really solving? Which assumption is weakest? What are we not seeing? Which competitor is reframing the category? Where is value leaking? What would make us change direction? What should we stop doing?

These questions are not slower. They are sharper. They prevent AI from becoming a speed layer on top of weak thinking.

Better questioning is not an individual trick. It is an organizational discipline. Leadership teams need to create environments where assumptions can be challenged without political cost, where uncertainty is made explicit, where AI outputs are interrogated, where teams distinguish facts from interpretations, and where decisions are connected to learning.

This requires a different meeting culture. Instead of starting with a deck, start with the decision. Instead of asking for updates, ask where reality differs from plan. Instead of asking whether the team is on track, ask what might make the current track wrong. Instead of asking how AI can make a task faster, ask whether the task should be redesigned. Instead of asking what the data says, ask what the data does not yet explain.

The best leaders will not use AI as a substitute for thinking. They will use it as a thinking partner, a challenger, a synthesizer and a pressure-testing mechanism. That means asking AI not only to produce, but to critique. Not only to answer, but to reveal uncertainty. Not only to summarize, but to compare interpretations. Not only to create options, but to expose trade-offs.

This is how AI improves executive judgment rather than replacing it.

A new gap is emerging inside organizations. Some teams will use AI to produce more. Others will use AI to think better. The first group will gain productivity. The second group will gain leverage.

The difference will be visible in the questions they ask. One team asks AI to write a campaign. Another asks which customer tension the campaign must resolve. One team asks for a sales script. Another asks which objections reveal a weakness in the proposition. One team asks for a market summary. Another asks which market signal could make the current strategy obsolete. One team asks AI to automate reporting. Another asks which decisions the report should improve and which metrics no longer matter.

This is the question gap. It will define the difference between AI activity and AI capability.

Organizations that close this gap will build better routines around problem framing, customer understanding, decision preparation and learning. Organizations that ignore it may still look busy with AI, but they will use it mainly to accelerate existing habits.

A useful way to assess AI readiness is to look at the questions circulating inside the organization. Are teams asking AI to produce more, or to think better? Are leaders asking for faster answers, or sharper decisions? Are AI outputs connected to action, or do they create more documents? Are assumptions being challenged, or simply made more polished?

AI will not save weak thinking. It may expose it. It may accelerate it. It may make it look more professional. But it will not automatically improve the quality of leadership judgment, strategic framing or executive decisions.

The organizations that win in the AI era will not be those that simply generate answers faster. They will be those that ask better questions earlier. They will know what decision they are trying to improve. They will challenge the assumptions inside the prompt. They will use AI to explore alternatives, not only confirm preferences. They will turn outputs into conversations, conversations into choices, and choices into execution.

In a world where answers are abundant, the scarce resource becomes question quality.
AI gets you from A to B faster. Better questions decide where to go.Imagination with AI takes you beyond the obvious path.

Harvard Business Review, The Surprising Power of Questions
Harvard Business Review, AI Prompt Engineering Isn’t the Future
Harvard Business Review, How Leaders Can Ask Better Questions
Harvard Business Review, Good Leadership Is About Asking Good Questions
MIT Sloan Executive Education, Why Asking Better Questions May Be the Most Important Leadership Skill in the AI Era
MIT Sloan Executive Education, The Leadership Skill AI Can’t Replace: Asking Better Questions
MIT Sloan, How Leaders Can Get the Most Out of Asking Questions
MIT Sloan Management Review, Apply AI Wisely in Decision-Making