When answers get cheap, good questions are the new scarcity
Standing out in a world where everyone has good answers
In 19th-century Paris, the Académie des Beaux-Arts defined what counted as legitimate art.
Realism, the prevailing standard, emphasized precision and visual accuracy. Success was based on how well you aligned with these norms. The system rewarded consistency, not experimentation.
The invention of photography in the 1830s and 1840s began to challenge this standard.
At first, photography seemed like a threat to painters. If a machine could record the world more precisely and more quickly than a human hand, what role did painting have?
But over time, photography freed painting from its representational obligations. Painters no longer had to compete with the camera in copying reality. Instead, they could focus on the subtleties that early cameras couldn’t capture. The play of light, the texture of perception. New interpretations of the familiar.
When I get to this point in the story, while speaking to an audience about AI and pause to see what they’re thinking, they typically respond with a knowing nod or a smile, because by that point, the parallel is obvious.
AI today is the camera. It won’t kill creativity or knowledge work. It will reinvent it.
Yes, but if that was all there was to it, this post would end right here.
There’s a bigger story behind the story. And in trying to get to simple lessons, we often miss out on that bigger story.
When answers become cheap…
The real story of the invention of the camera is the story of what happened next to the world of art.
Without photography, art would have progressed - at least for some more time - on a predictable trajectory towards more of the same. More Realism - improvements in accuracy.
If Realism was the prevailing answer of the time, artists would have gone on to give better answers.
Photography, ironically, collapsed the cost of generating answers. You could get the most realistic portraits without hours of effort on the part of the artist.
Photography freed painting from Realism, but what was most interesting was what simultaneously rose to take its place.
Impressionsts like Monet and Degas began experimenting with subjective experiences of color and light. Instead of representing reality, which the camera could do with far less effort, they started interpreting it.
Instead of providing better answers - more Realism - the Impressionists were redefining the question altogether.
With Realism, art was judged based on its ability at representation.
With Impressionism, art had a new purpose: interpretation.
The camera provided cheap replicas - abundant answers.
The Impressionists decided to change the framework - and position art as a basis for asking better questions.
When what was previously scarce suddenly becomes abundant,
look for the new scarcity.
Because that’s what creates leverage.
LLMs and cheap answers
LLMs are the latest step in a long arc of technologies that have made answers progressively cheaper.
From data analytics to recommendation engines to ChatGPT, each wave has broadened access and driven the cost of answers toward zero.
These systems specialize in answers. Not necessarily correct ones, and rarely final ones, but answers that are immediate, and most importantly, plausible.
And as far as LLMs are concerned, answered with confidence!
The problem, though, is that
plausible answers are worse than those that are clearly wrong!
When an answer feels good enough, we tend to stop asking.
In an environment overloaded with content and starved for attention, plausibility becomes a stand-in for truth.
Search results that confirm our assumptions rise to the top, memes that reinforce ideas in our head get shared further, language models that complete our thoughts reinforce existing narratives.
The cost of continuing the inquiry rises.
Good questions become more expensive than ever.
Alongside this, the world around us is becoming less stable.
As I’ve noted before, we’re moving into a world of structural uncertainty - one where the rules are no longer static.
What worked yesterday may not apply tomorrow, not because the facts have changed, but because the terrain has.
In such an environment, answers that were once reliable quickly become outdated. Static knowledge has limited utility in dynamic systems.
What matters more is the capacity to stay curious,
and continue a line of inquiry.
This is where good questions - even though expensive - become strategic. A good question expands the field of awareness. It reframes the problem.
In systems marked by structural uncertainty, value is created not by declaring what is known, but by directing attention to what remains unresolved.
The most valuable answers today are not those that appear most complete and articulate, but those that reveal where we must continue looking.
In a system with structural uncertainty, the goal is no longer certainty, but continued navigation, to constantly find orientation in a moving landscape.
A good question that framed the future
Much of the world we live in today can be traced back to a good question that was asked in 1948.
Back then, engineers at Bell Labs were trying to improve the clarity of phone calls. They were looking for better answers by tinkering with wires, amplifiers, and filters.
Claude Shannon, though, was asking a different question.
Instead of asking how to reduce noise on telephone lines, he asked something far more fundamental:
What is information?
This question, rather, this detour into abstraction, gave birth to information theory, a body of work that defined how much uncertainty a message could contain.
Shannon’s key insight was that the value of information (answers)
is proportional to the uncertainty it resolves.
What is a good answer?
According to Shannon:
The value of an answer is in the uncertainty it resolves - and hence becomes actionable.
A good way to understand this is to think about a weather forecast. A forecast that tells you it will be sunny in the middle of summer might sound fine, but it doesn’t change your behavior. You already assumed it would be sunny. But a forecast that warns of unexpected rain, that’s a lot more useful. That changes your decision, you bring an umbrella.
Information theory is built around the idea of managing uncertainty. Not around the idea of providing plausible-sounding answers.
But the dominant systems we’ve constructed - LLMs being the latest - do the opposite. They dazzle us with their fluent, convincing answers.
This is the first trap.
Cheap and easily available answers only serve to reduce attention (which is already scarce), they do not reduce uncertainty.
LLMs produce language that sounds authoritative but often offers little in the way of new insight. The better they get at sounding right, the easier it becomes to stop looking further.
And when answers are cheap and abundant, the process of inquiry becomes expensive. The very act of asking slows you down, not because the questions have been answered, but because you’re led to believe that they have.
For a generation trained (conditioned, hooked, beahvior-designed) to doom-scroll and binge-watch, the allure of cheap and abundant answers is undeniable.
And the cost of asking the right questions is way too high - like pulling yourself out to the gym one week out from your New Year’s resolution.
But as Shannon understood, not all answers are created equal.
The second trap is believing that more answers are better than fewer answers.
Not really.
In knowledge-rich, attention-poor environments, the problem isn’t scarcity of facts, it’s the misallocation of attention.
We continue to gather data long after its marginal utility has declined, because the systems we’ve built are designed to answer, not to question.
They’re set up for speed and verbosity (try having a conversation with Claude), not for surfacing what’s ambiguous or incomplete.
A good answer, then, is one that
(1) reduces uncertainty and
(2) values limited attention well so as not to misuse it.
The power of good questions in a world of cheap answers
Answers often live within the constraints of existing frames.
Questions, when they are powerful, widen the frame itself.
The biggest trap in operating with traditional frames is that we end up answering questions of structural uncertainty using operational frames and tactics.
Copernicus and Einstein - or more recently, CRISPR, which sat in scientific literature for two decades as a bacterial immune system before someone changed frame and applied it as a tool for gene editing - weren’t answers to existing questions. They were questions that exposed the limitations of prevailing assumptions.
Once the frame shifted, the uncertainty resolved.
Pre-order my upcoming book at 70% off
This post is based on ideas from my upcoming book Reshuffle.
Reshuffle is now available for pre-orders. All pre-orders leading up to the launch date are at 70% off.
Pre-orders are Kindle only.
Hardcover, paperback, and audiobook versions will be available at launch.
The surprising human + LLM advantage
Finally, structural uncertainty is often resolved not in following the most obvious path, but in the unexpected intersection between distant ideas. This is where good questions become critical, because only a question can bridge concepts that don’t already belong to the same system.
This is one place where a human asking good questions can get superpowers when paired an LLM throwing out cheap answers. LLMs are particularly good at connecting unconnected domains, but only if they’re asked the right questions.
In the hands of someone looking for cheap answers, LLMs are a liability.
But in the hands of those asking good questions, LLMs could be superpowers.
Good questions change the world
Shannon’s original question, posed in the late 1940s with no particular commercial urgency, turned into a seed for an entirely new way of engaging with complexity. His work introduced a new logic for understanding systems under uncertainty.
The first-order impact was immediate. Shannon made information measurable - quantified in bits. With that, engineers had instruments to calculate uncertainty, allocate bandwidth, and measure how much information could safely be transmitted through noise.
The second-order effects were more transformative. Shannon’s framework allowed communication systems to be built not just to carry information, but to protect and compress it. Compression algorithms, cryptographic protocols, memory storage - all grounded in Shannon’s ideas - gave us the internet and digital media.
But the third-order effects were far greater, and cultural. Biologists began to describe DNA as a code. Cognitive scientists framed thought as an information process. Economists spoke of information asymmetries. Physicists speculated that the universe might be, at some level, made of information.
The question had escaped beyond the bounds of engineering.
All of this began with a single reframing: What is information?
Instead of looking for cheaper answers,
Shannon built a system that helped the world ask better questions.
The irony of our current moment is that the more knowledge we accumulate, the less certain we seem to become.
When plausible answers come faster than we can formulate questions, the challenge lies in structuring the inquiry, and knowing where to look next.
Which paths of exploration still hide useful ambiguity.
In the midst of structural uncertainty i.e. uncertainty not just in outcomes but in the structure of the systems which deliver those outcomes, advantage lies not in what you know, but in your ability to navigate what you don’t.
When answers aren’t static anymore, the ability to keep asking the right questions is the only thing that matters.
The consulting conundrum
There’s a little something that I like to call the consulting conundrum.
In a world flooded with highly plausible answers, it’s increasingly easy to fall into traps that don’t feel like traps at all initially.
As a client, you might assume that answers are easy to come by.
And because answers seem abundant, you start evaluating them the way you’d evaluate any commodity: by cost.
But that logic leads to a race to the bottom. The cheaper the answers get, the more likely it is that no one has done the expensive work of asking the right questions. You save money upfront, but what you get in return is almost always useless, or worse still, misleading.
On the other hand, if you bring in someone else to help you navigate uncertainty, you’re left navigating another trap.
You’re now paying for guidance, but unless you know how to assess not just the answers, but - far more importantly - the quality of the questions being asked, you have no reliable way to judge the value of what you’re buying.
Confidence is not a substitute for clarity.
Style can be used to dress up, what Richard Rumelt would call, bad strategy.
That’s the conundrum.
The threat to consulting isn’t that AI will replace firms or automate junior work.
The real issue is that the economics of questions and answers are changing.
In a world where good answers are increasingly easy to generate, the bottleneck shifts to the framing of the problem itself.
Unless clients learn how to recognize the economics of good questions, they’ll find themselves caught in one of two traps.
Either they treat answers like a commodity and get exactly what that model rewards: low-cost, low-value, lowest-common-denominator thinking.
Or they pay for confidence and showmanship, without knowing whether they ever actually solved the right problem.
To navigate this shift, clients need to develop a new muscle: the ability to evaluate thinking not by fluency or showmanship, but by the structure of the inquiry that led to it.
In a world of cheap answers, the most expensive mistake you can make is confusing cheap answers as good answers.
The re-skilling cop-out
The cop-out answer to AI is re-skilling. Whenever someone talks about AI and jobs, pat comes the answer - we will have to invest in re-skilling displaced workers.
But if re-skilling only helps you learn new answers, that’s not very useful. It’s only a matter of time before AI improves towards making those answers cheaper.
The most valuable capability lies in improving our capacity to frame better questions.
This is a very different skill than we are taught to value in today’s education and skilling system. For most of the twentieth century, success came from mastering a domain or building expertise. We were tested on our ability to give answers, not on our ability to ask questions. More knowledge was better. But this logic of accumulation only works if the rules don’t change.
When conditions are stable, answers compound.
When conditions shift, constant inquiry matters more.
What matters now is learning how to move through complexity without getting stuck in the illusion that we already understand it.
The best knowledge workers will treat uncertainty not as a threat to be managed but as a terrain to be explored. They’ll look to construct partially but directionally correct maps, instead of falling for the trap of certain answers.
Want to get early access to Reshuffle as a beta reader?
I’m looking to bring on a cohort of beta readers to read the book and provide feedback. The book will be opened out to this cohort between April 15 - May 15.
If you’d like to join this group, please click below. Please ensure that you have sufficient bandwidth during this period to provide feedback on the content of the book.
Good piece. I spent Saturday conducting a quarterly business review with one of my companies, exploring the right questions to ask. One particular question took us far away from the current business but solved an even bigger problem and positioned the business really well.
I think at lot of the time we don't think "big or wide" enough with our questions.
One of the better thoughts i have read recently. To be or not to be is the question