Claude Shannon: The Man Who Never Stopped Playing

Why curiosity, uncertainty, and human judgment matter more than ever in an information-overloaded world.

People who see something the rest of us miss fascinate me. Not necessarily because they are smarter. Sometimes they are simply more curious.

Claude Shannon was one of those people.

If you use a computer, send a text, watch a video, make a phone call, search the internet, or use AI, you are living in a world shaped, in part, by the way Shannon thought about information.

That alone would make him worth writing about. But that is not really what caught my attention. What caught my attention was the way he worked.

Shannon was a serious mathematician and engineer, but he apparently never believed serious work had to look serious. He juggled. He rode a unicycle through Bell Labs. He built strange machines because he wanted to see if they would work. He played with chess, puzzles, mechanical devices, and ideas that did not always have an obvious commercial purpose.

And somewhere in all that playing, he helped create the foundation for the digital world.

I love that.

Because we spend so much time now asking whether something is useful. What is the ROI? Can it generate business? Can we turn it into content? Can AI do it faster? Can it save us time?

All reasonable questions. I ask them too. But I wonder if we have become so obsessed with efficiency that we are in danger of squeezing curiosity out of the process.

Some of the best ideas do not start with a business plan. They start with "I wonder what would happen if..."

That may be one of the most valuable sentences in business. And maybe in life.

Information is supposed to reduce uncertainty

There is an observation widely attributed to Shannon that I find particularly relevant today:

"Information is the resolution of uncertainty."

Think about that for a moment.

Shannon was dealing with information mathematically, measuring uncertainty in communication systems. He was not trying to explain social media, politics, or the daily avalanche of opinions we encounter. But I think the idea has a much broader lesson for us.

We have more information at our fingertips than any generation in history. We can ask AI almost anything and receive a beautifully organized answer in seconds. We can read ten different interpretations of the same event before breakfast. We can watch experts disagree about the economy, technology, healthcare, business, and just about anything else.

And after consuming all that information, how often do we actually feel more certain? Sometimes we are more confused than when we started.

That strikes me as one of the great contradictions of our time.

We have become extraordinarily good at producing information and remarkably less certain about what to do with it.

An answer can sound authoritative without being accurate. A confident prediction can be completely wrong. An opinion repeated a thousand times does not magically become a fact. And an AI-generated response can be beautifully written and still miss the point.

The problem is no longer finding information. The problem is knowing what deserves our attention, what we should believe, and what we can actually use.

We are living in a strange moment

We now have tools that can give us answers faster than most of us ever imagined. I think that is wonderful. I use them. I experiment with them. I discover new things I would not have thought to investigate otherwise.

But there is something about this rush toward instant answers that bothers me. We seem to be placing enormous value on how quickly we can arrive at a conclusion. I'm not convinced speed is always our friend.

Consider what is happening in business today. A company can use AI to analyze its market, study its competitors, generate a strategy, create presentations, and recommend its next move.

Impressive. But what happens when the market behaves differently than predicted? When a customer does something unexpected? When the numbers tell one story and experience suggests another?

Someone still has to decide what to believe and what to do. And that decision requires something more than the ability to retrieve information. It requires judgment.

After decades in business, I have learned that some of the most important decisions are made when the available information is incomplete, contradictory, or simply wrong. You listen. You question. You look for what might be missing. And sometimes you make a decision that the numbers alone would never have suggested.

That is not an argument against AI or analytics. Quite the opposite. It is an argument for understanding what they can do and recognizing what remains our responsibility.

When answers are easy, questions become more valuable

This is where Shannon's curiosity becomes particularly interesting to me.

He was not simply looking for better answers. He was willing to ask questions other people might have dismissed as pointless.

What if a machine could do this? What if information could be measured? What if something that appeared random followed a pattern? What if an idea that seemed impractical today turned out to be important tomorrow?

I think we need more of that thinking.

AI is becoming exceptionally good at helping us explore what is already known. But we still need people willing to challenge assumptions, notice contradictions, and ask questions that take us somewhere unexpected.

Suppose two consultants ask AI how to improve their businesses. Both receive reasonable recommendations. Both use the same technology. Both have access to much of the same information.

What makes one more successful than the other?

It may be that one accepts the recommendations while the other asks, "What are we missing?" Or, "Why are we doing it this way in the first place?"

Or, my personal favorite: "What happens if we try something entirely different?"

That is where things get interesting.

More content does not necessarily mean more clarity

I spend much of my time working with people who have accumulated years of knowledge, experience, ideas, and insights. And I see an interesting problem developing.

We have made creating content easier than ever. A single idea can become a dozen LinkedIn posts, a newsletter, a video, a podcast, a presentation, and a course.

Wonderful. But suppose all those pieces simply repeat the same general observations everyone else is making. What have we accomplished?

More content? Certainly. More information? Perhaps. More understanding? Not necessarily.

I believe the real value of someone's expertise lies in their ability to help another person understand something they did not understand before. To see a problem differently. To recognize an opportunity. To make a better decision. To feel a little less uncertain about what comes next.

That is where Shannon's observation resonates with me.

The value of what we communicate should not be measured only by how much we produce, but by how much clearer we make something for someone else.

That is not Shannon's mathematical definition of information. It is the practical lesson I take from it. And I think it matters enormously in an AI-first world.

Sometimes the most productive thing you can do is play

There is something else I like about Shannon. He did not seem overly concerned with looking productive.

That may sound like a small thing, but I do not think it is.

We have turned being busy into something of a competitive sport. Every hour has to produce something. Every project needs an outcome. Every idea needs a destination. Now we have AI promising to make all of it happen even faster.

I have nothing against efficiency. I've spent a good part of my career trying to improve it. But I also know that some ideas need room to wander before they become useful.

Sometimes you have to follow the wrong path before you find the right one. Sometimes you discover something important while working on something entirely unrelated. And sometimes the thing that looks like wasting time turns out to be the most valuable thing you did all week.

After a long time in business, I have become much more comfortable with that. You do not always have to know where an idea is going before you follow it.

In fact, knowing exactly where you are going can sometimes prevent you from discovering somewhere better.

What Shannon might teach us about the future

I have no idea what Claude Shannon would think of today's AI systems. I suspect he would be fascinated.

I also suspect he would take one apart, figure out how it worked, and then try to make it do something nobody had thought of. Probably while riding a unicycle.

But the more I think about his life, the more I believe there are two lessons worth carrying into this extraordinary period.

The first is that information has value when it helps resolve uncertainty. The second is that curiosity is how we discover what we do not yet know.

We need both.

We need technology that helps us process the overwhelming amount of information surrounding us. And we need human beings who are willing to question that information, explore unfamiliar ideas, and exercise judgment when the answers are not obvious.

I don't believe our future depends on choosing between human intelligence and artificial intelligence. I believe it depends on what happens when we use both intelligently.

Technology will continue to get faster. AI will continue to improve. Information will become even easier to create. But I hope we never become so efficient that we forget how to be curious.

Because the most interesting discoveries rarely begin with someone announcing they already know the answer. They begin with someone who isn't afraid to admit they don't. And who finds that uncertainty interesting enough to explore.

So I'll leave you with two questions.

When was the last time you followed an idea simply because you were curious where it might take you?

And with all the information available to us today, are we becoming better informed, or simply better supplied with answers?