Imagine two investors.

The first puts his entire retirement savings into a single speculative investment. He hasn’t researched it particularly well, but someone he trusts tells him it’s going to be the next big thing.

Six months later, the investment doubles.

The second investor spends considerable time researching an opportunity. She examines the business, considers the risks, and determines how much she can afford to lose. She invests a modest portion of her portfolio.

Six months later, the investment falls by 30%.

Who made the better decision?

Most people would instinctively point to the first investor. After all, he doubled his money while the second investor lost nearly a third of hers.

But what if the first investor had a substantial chance of losing everything?

And what if the second investor had made a reasonable assessment of the opportunity, understood the downside, and invested an amount she could comfortably afford to lose?

Would we still consider the first decision better?

This illustrates one of the most persistent mistakes in how we evaluate decisions.

We confuse the quality of a decision with the quality of its outcome.

And that confusion can be expensive.

The Problem With Judging Decisions Backwards

When we know how something turned out, it’s remarkably difficult to evaluate the original decision without allowing that knowledge to influence our judgment.

Psychologists call this outcome bias.

In a well-known 1988 study, researchers Jonathan Baron and John Hershey found that people rated decisions more favourably when they were told those decisions had produced successful outcomes, even when the information available to the decision-maker was otherwise identical.

In other words, knowing the result changed how people evaluated the decision itself.

Think about how often we see this in everyday life.

An entrepreneur takes an enormous financial risk, builds a successful company, and becomes celebrated for having the courage to bet everything.

Another entrepreneur makes a similar bet, loses everything, and is criticised for being reckless.

Yet the original decisions may have been almost identical.

One happened to succeed. The other didn’t.

We see the same phenomenon in investing.

Someone buys a stock because a friend recommended it. The stock rises, and the investor concludes that trusting the recommendation was a good strategy.

Someone else carefully evaluates a business, buys its shares at what appears to be a reasonable valuation, and subsequently loses money because of an unexpected development.

The first investor feels clever. The second questions their judgment.

But the outcomes tell us much less about the quality of their decisions than we might think.

Of course, outcomes aren’t irrelevant.

Over time, the results of repeated decisions can tell us a great deal about whether our approach is working.

But a single outcome is a particularly unreliable measure of decision quality.

A good process can produce a bad result. A bad process can produce a good result.

The danger comes when we mistake the latter for evidence that our process was sound.

Success Can Be a Terrible Teacher

Failure has an obvious disadvantage. It hurts.

But it also has one useful characteristic: it tends to make us question what we did.

Success doesn’t always have that effect.

When something works, we’re inclined to repeat it. Sometimes that’s sensible. Other times, we’re simply reinforcing a behaviour that happened to produce a favourable outcome.

Consider someone who invests heavily in a rapidly rising market.

Their investments perform exceptionally well for several years.

They conclude that their strategy is sound, perhaps even that they have a particular talent for identifying opportunities.

So they increase their exposure.

But what if much of their success came from favourable market conditions rather than superior judgment?

The better their investments perform, the more confident they become.

And the more confident they become, the greater the risks they’re willing to take.

The strategy may appear increasingly successful right up until the point when the underlying conditions change.

This is why a long period of success doesn’t necessarily demonstrate that someone has been managing risk well.

Sometimes it merely demonstrates that the risks haven’t materialised yet.

Seeing the difference is especially important because success can often encourage us to increase the size of our commitments.

A mistake involving $1,000 is one thing.

Making the same mistake with $1 million because the first ten attempts happened to work is quite another.

Success can teach us what worked. It doesn’t necessarily teach us why it worked.

And unless we understand the difference, we may be learning precisely the wrong lesson.

Probability Is Only Half the Risk Equation

When people talk about risk, they often focus on the probability of something going wrong.

What’s the likelihood that this investment will lose money?

What’s the probability that this business will fail?

How likely is it that I’ll need to work longer than planned?

These are sensible questions.

But they’re incomplete.

Because the probability of an adverse outcome tells us nothing about whether we can afford the consequences.

Imagine being offered an opportunity with a 90% chance of doubling your money and a 10% chance of losing everything you invest.

On the face of it, those odds look attractive.

For every $100 invested, the expected ending value is $180: a 90% chance of receiving $200 and a 10% chance of receiving nothing.

That’s an expected gain of 80%.

But would you invest your entire retirement savings?

I wouldn’t.

Not because the odds are unattractive, but because the consequences of being wrong could be devastating.

A 10% probability of losing everything is not a small risk when everything is what you stand to lose.

Now imagine the same opportunity, but instead of committing your entire retirement savings, you invest 1% of your portfolio.

The probability of failure hasn’t changed.

The investment hasn’t become intrinsically safer.

But the consequences of failure have changed dramatically.

You could lose the entire investment and still have 99% of your portfolio remaining, before accounting for any other changes in its value.

This illustrates an important principle:

Risk isn’t just the probability of being wrong. It’s also the consequence of being wrong.

And the consequence depends not only on the investment or decision itself, but also on your personal circumstances.

A $50,000 loss might be a manageable setback for one person and a financial catastrophe for another.

The same opportunity can therefore be reasonable for one person and completely inappropriate for someone else.

That is why I have always been uncomfortable with investment discussions that focus primarily on potential returns.

Returns matter.

But so does the ability to withstand the outcomes we don’t want.

The Hidden Cost of Being Unable to Recover

There’s another aspect of risk that deserves more attention.

Some losses are recoverable.

Others permanently change what we’re able to do next.

Suppose you have $100,000 invested.

If your portfolio loses 10%, you’re left with $90,000. You need an 11.1% gain to recover your original capital.

If you lose 50%, you’re left with $50,000. You now need a 100% gain to get back to where you started.

And if you lose 80%, you need a 400% gain just to break even.

Loss	Gain needed to recover
10%	11.1%
25%	33.3%
50%	100%
80%	400%
90%	900%

This mathematical asymmetry has enormous implications.

As losses become larger, recovery becomes disproportionately more difficult.

And that’s before considering the time required to recover.

For someone in their 30s, a major financial setback might leave decades to rebuild.

For someone approaching retirement, the same setback could permanently alter their choices.

This doesn’t mean older investors should avoid all risk.

In fact, avoiding risk entirely introduces other risks, including the possibility that inflation will erode purchasing power or that investments won’t generate sufficient returns.

But it does mean that the capacity to recover from a loss should influence how much risk we’re prepared to take.

This is one reason I believe the ability to remain financially viable is so important.

You only compound what you keep.

An investment strategy that produces spectacular returns for several years but eventually wipes out your capital doesn’t allow you to enjoy the benefits of long-term compounding.

The same principle applies beyond investing.

A business decision that could destroy an otherwise healthy company may be unacceptable even when the probability of success appears attractive.

A career decision that sacrifices financial security for an uncertain opportunity may have consequences extending far beyond the immediate loss of income.

The question isn’t simply whether the decision might work.

It’s whether you can live with what happens if it doesn’t.

Good Decisions Preserve Future Choices

This brings us to something I think is frequently overlooked.

When evaluating a decision, we tend to compare the immediate outcomes of the available alternatives.

Which investment offers the greatest return?

Which business opportunity has the highest potential?

Which career move offers the largest salary increase?

But there’s another consideration.

What choices will I still have after making this decision?

Imagine two people who want to start a business.

The first resigns from a well-paying job, invests a substantial portion of their savings, and spends a year building a product before discovering whether customers want it.

The second keeps their job, identifies a problem worth solving, develops a modest initial solution, and tests it with potential customers before committing significant resources.

Neither approach guarantees success.

The first entrepreneur might build a hugely successful business.

The second might discover that nobody wants the proposed solution.

But notice what happens if both are wrong.

The first may have lost a year of income, a substantial amount of capital, and perhaps the financial flexibility to attempt something else.

The second may have lost a relatively small amount of time and money, while gaining useful information about the market.

And because the second entrepreneur has preserved more of their resources, they remain in a position to try something different.

This is the value of preserving optionality: retaining the ability to respond to new information and changing circumstances.

It’s particularly important when decisions involve substantial uncertainty.

If you can test an assumption cheaply before making an expensive commitment, why wouldn’t you?

If you can explore an opportunity without surrendering your existing security, why make the exploration dependent on a dramatic leap?

And if you can structure a decision so that an unfavourable outcome leaves you with useful information rather than permanent damage, isn’t that worth considering?

Sometimes the best decision isn’t the one with the greatest immediate upside.

It’s the one that gives you a reasonable opportunity to benefit while preserving your ability to make good decisions tomorrow.

But Preserving Options Has a Cost Too

There’s an important qualification here.

If we take the idea of preserving options too far, we can end up doing nothing.

We avoid making commitments because we want to keep every possibility available.

We postpone investments because markets might fall.

We delay starting businesses because our ideas might fail.

We remain in unsatisfying careers because changing direction involves uncertainty.

And we tell ourselves that we’re being prudent.

But indecision isn’t necessarily risk-free.

Keeping money entirely in cash can expose us to inflation.

Avoiding investments can mean missing years of potential compounding.

Postponing a business idea indefinitely can mean losing the opportunity to learn, build relationships, or establish something valuable.

Even maintaining the status quo is a decision, whether we acknowledge it or not.

So the objective cannot simply be to minimise the possibility of loss.

If that were the case, the safest strategy would often appear to be doing nothing.

And doing nothing can carry substantial long-term costs.

The more useful objective is to take risks deliberately, in proportion to what we can afford to lose and what we stand to gain.

That requires judgment.

Not the elimination of uncertainty, but an understanding of which uncertainties are worth accepting.

How Should We Judge a Decision?

If outcomes alone aren’t sufficient, what should we use instead?

I think there are four questions worth asking.

First, what did I reasonably know at the time?

A decision should be evaluated using the information that was available when it was made, not information that became apparent afterwards.

That doesn’t excuse ignoring evidence or failing to investigate something important.

But it does recognise that hindsight gives us knowledge the original decision-maker didn’t possess.

Second, were my assumptions reasonable?

Every meaningful decision involves assumptions.

The problem isn’t that we make them. It’s that we sometimes forget we’re making them.

A retirement plan might assume a particular investment return.

A business plan might assume customers will pay a certain price.

A career decision might assume a particular industry will continue growing.

Good decision-making requires us to identify those assumptions and consider what happens if they prove wrong.

Third, could I afford the downside?

Not merely in financial terms, but in terms of time, health, relationships, responsibilities, and future opportunities.

A risk that is perfectly acceptable in one set of circumstances might be unacceptable in another.

Fourth, what would I learn regardless of the outcome?

This is especially useful when dealing with uncertainty.

Some decisions can be structured as experiments.

Instead of committing everything to an assumption, we can make a smaller commitment that produces information.

If the results are encouraging, we can commit more.

If they aren’t, we can reconsider.

The aim isn’t to avoid being wrong.

It’s to make being wrong less costly, while increasing our understanding of what might work.

The Real Measure of a Good Decision

Over the years, I’ve sat on boards and investment and risk committees where decisions involved substantial amounts of money, competing priorities, and uncertain outcomes.

One thing that experience reinforced is that even careful analysis cannot eliminate uncertainty.

You can examine the numbers.

Challenge assumptions.

Consider alternative scenarios.

Seek expert opinions.

And still arrive at a decision that doesn’t produce the outcome you hoped for.

That doesn’t necessarily mean the decision was poor.

Nor does a favourable result prove that the original reasoning was sound.

What matters is whether the decision was justified by the information available, whether the risks were understood, and whether the consequences were acceptable.

And, importantly, whether the decision left sufficient capacity to respond when circumstances changed.

Because life doesn’t give us just one decision to make.

We make thousands of them.

Some work. Some don’t.

Some produce unexpected opportunities. Others produce disappointments.

And occasionally, something that initially looks like a failure turns out to be the beginning of something better.

We cannot control every outcome.

But we can improve how we make decisions, how much we commit to them, and how we respond when reality differs from our expectations.

Perhaps that’s a better way to think about success.

Not as a perfect record of correct predictions.

Not as a life in which nothing ever goes wrong.

But as the cumulative result of decisions that allow us to learn, adapt, recover, and continue moving forward.

You don’t have to be right every time to build a successful future.

You need to make decisions that allow you to survive being wrong—and benefit when you’re right.

There’s a peculiar contradiction in the way we make decisions today.

We have access to more information than any generation in history.

We can compare investment strategies, calculate retirement scenarios, research business opportunities and ask AI to explain almost anything we don’t understand.

Questions that once required a specialist, a library or several days of research can now be answered in seconds.

You would expect all this information to make us more confident about our decisions.

Yet many people seem to be experiencing the opposite.

They know more.

They have more options.

They can see more possibilities.

And somehow, they find it harder to decide what to do.

Perhaps the problem isn’t that we don’t have enough information.

Perhaps we’ve misunderstood what information is supposed to do for us.

The retirement confidence paradox

A recent study by Singapore wealth management platform Endowus illustrates the problem rather well.

Its 2026 Wealth Insights Report surveyed 500 Singaporeans about retirement planning, financial advice and AI.

The findings were striking.

75% considered a secure retirement important. Yet only 28% felt confident that their plans were on track.

In other words, almost three-quarters weren’t confident they were heading toward the retirement they wanted.

And it wasn’t simply because they didn’t know retirement planning was important.

They knew.

Many had already taken steps.

But they remained uncertain about whether those steps would get them where they wanted to go.

The report also examined how many information sources people used when making financial decisions.

Those who felt confident used an average of 2.58 sources.

Those who weren’t confident used 2.43.

That’s hardly the dramatic difference we might expect if access to more information were the principal source of confidence.

The more revealing finding concerned planning. Respondents with a clear plan they were executing were substantially more likely to feel confident about their retirement prospects than those who had taken some steps but weren’t sure whether they were on track.

Now, a survey about confidence doesn’t establish whether someone’s retirement plan will actually succeed.

People can feel confident and be wrong.

They can also feel uncertain despite having made perfectly sensible decisions.

But the findings raise an interesting question.

If information is so readily available, why is it still so difficult to turn what we know into decisions we can act on?

Information and understanding aren’t the same thing

Imagine you’re trying to decide whether you can afford to retire in five years.

You begin researching.

You discover articles about how much money people supposedly need for retirement.

You read about the 4% withdrawal rule.

You compare dividend-paying investments, bonds, annuities and property.

You calculate how inflation might affect your expenses.

Then you ask AI to evaluate your retirement portfolio.

It produces an impressively detailed analysis, complete with projections, assumptions and possible strategies.

You’ve learned an enormous amount.

But can you retire in five years?

You still don’t know.

And the reason isn’t necessarily that you need more information. It’s that the information you’ve collected hasn’t yet been organised into an understanding of your particular situation.

A retirement calculation that assumes you’ll live to 85 may look perfectly reasonable.

But what if you live to 95?

A portfolio projection that assumes a 7% annual return may appear reassuring.

But what happens if the market falls 30% shortly after you retire?

You can gather more information about longevity, historical returns and market volatility.

Yet none of it will tell you precisely how long you’ll live or what markets will do.

The information may help you understand the risks.

It cannot eliminate them.

So you can continue researching indefinitely, hoping the next piece of information will finally make the decision obvious.

Or you can recognise that the problem has changed.

You no longer need to answer every possible question about retirement.

You need to identify which uncertainties matter to your decision, which you can do something about, and which you must learn to manage.

That’s a different kind of work.

More information can create more uncertainty

We tend to assume that uncertainty comes from not knowing enough.

Sometimes it does.

If you’re buying a house and don’t know whether the roof is leaking, a building inspection can reduce uncertainty.

If you’re considering an investment and don’t understand how it generates returns, learning about its business model can help.

Information is extraordinarily useful when it answers a question that matters.

But not every additional piece of information makes a decision easier.

Suppose you’re considering changing careers.

Initially, you see two options: remain in your current job or find another.

Then you begin researching.

You discover consulting, freelancing, starting a business, retraining, investing for income, part-time work and several other possibilities.

AI can help you explore every one of them.

Within an afternoon, you might have ten plausible alternatives, each with its own advantages, disadvantages, risks and requirements.

You now know considerably more than when you started. But you also have more decisions to make. And every new option creates additional questions.

What if consulting is better than employment?

What if starting a business offers greater long-term potential?

What if retraining makes more sense?

What if AI changes the economics of all three?

The problem is no longer a shortage of possibilities.

It’s that you haven’t established how to choose between them.

Information has expanded your choices without giving you a basis for making the choice.

Clarity begins by removing what doesn’t matter

Consider two people evaluating the same investment.

Both have access to identical information.

The first person wants the highest possible return.

The second needs reliable income to meet living expenses over the next ten years.

An investment offering substantial growth potential but unpredictable cash flow might be attractive to the first person and unsuitable for the second.

Neither person necessarily has better information.

They have different purposes.

And those purposes change which information matters.

This is why I believe clarity begins before analysis. You first need to understand what you’re actually trying to achieve. Only then can you determine which information is relevant.

Otherwise, you risk spending enormous amounts of time analysing things that have little bearing on your decision.

It’s rather like planning a journey.

You can study every road, train route and airline schedule in the country.

You can compare prices, travel times and reliability.

But until you’ve decided where you’re going, you cannot determine which route is best.

The information isn’t useless.

You simply haven’t established what you’re using it for.

The same principle applies to careers, businesses, investments and retirement.

Before asking which option is best, ask:

Best for what?

That question can eliminate an extraordinary amount of unnecessary complexity.

Understanding also means knowing what you might be getting wrong

Even when we know what we’re trying to achieve, there’s another obstacle.

Our conclusions depend on the assumptions we make.

And assumptions are particularly dangerous because they often don’t look like assumptions.

They look like facts.

Someone might assume that because they’ve earned a high salary throughout their career, they’ll always be able to find another well-paying job.

Someone else might assume that because an investment has performed well for the past decade, it will continue performing well.

A third person might assume that because they’ve never run a business, they wouldn’t be capable of building one.

These beliefs can shape decisions long before anyone thinks to question them.

You can conduct extensive research and still arrive at a poor conclusion if the research is built around the wrong assumption.

In fact, the additional information may make you more confident in that conclusion.

Because the problem isn’t always what you don’t know.

Sometimes it’s what you think you know.

This is where judgment becomes important.

Judgment involves recognising which assumptions deserve scrutiny, which risks are significant and which details are unlikely to change the outcome.

It’s the ability to distinguish between a question that genuinely needs answering and one that’s merely interesting.

And that ability isn’t necessarily improved by collecting more information.

It improves when you learn to examine how you’re interpreting the information you already have.

A good decision doesn’t require a certain outcome

There’s another reason people can remain uncertain despite extensive research.

They may be asking information to provide something it cannot.

A guarantee.

Think about an important decision you’ve made in your life.

Perhaps accepting a job.

Buying a property.

Getting married.

Making an investment.

Leaving a career.

At the time, you couldn’t possibly have known everything that would happen afterward.

You made the decision using the information available, your understanding of the situation, and your judgment about the risks and opportunities.

Some decisions worked out well.

Others probably didn’t.

But an unfavourable outcome doesn’t automatically mean the original decision was poor.

You might make a sensible investment that loses money because of an unforeseen event.

Or a reckless investment that makes money because you happened to get lucky.

If we judge decisions entirely by their outcomes, we risk learning the wrong lessons from both.

A better question is whether the decision was reasonable given what you knew, what you could reasonably have discovered, and the alternatives available at the time.

Did you understand what you were trying to achieve?

Did you identify the important assumptions?

Did you consider what could go wrong?

Could you afford the consequences if things didn’t work out?

Were there ways to test the idea or limit your exposure before making a larger commitment?

Those questions won’t guarantee success.

But they can help you make decisions without needing to predict the future.

Clarity is not certainty about what will happen. It’s understanding enough to make a decision despite not knowing what will happen.

The danger of waiting until you know enough

Of course, there’s a reasonable objection to all this.

Surely it’s better to research carefully than to make an uninformed decision?

Absolutely.

The problem isn’t research.

It’s failing to recognise when additional research is no longer improving the quality of the decision.

Imagine someone who has been thinking about starting a small business for three years.

They’ve read books, watched videos, attended webinars and researched dozens of business models.

They understand the terminology.

They can explain the advantages and disadvantages of different approaches.

But they’ve never spoken to a potential customer.

They’ve never tested whether anyone would pay for what they want to offer.

They’ve never taken the smallest practical step toward finding out whether their idea works.

At some point, another course or another book is unlikely to resolve their uncertainty.

Because the missing information can only be obtained through action.

They don’t need another explanation of how businesses work.

They need evidence about whether their particular idea has a market.

The same applies to many other decisions.

You can research a career change for months, but a conversation with someone already doing the work may reveal more than another hundred articles.

You can study investment strategies indefinitely, but until you’ve established your actual income requirements and risk tolerance, you may still be unable to choose.

Sometimes understanding comes from analysis.

Sometimes it comes from experience.

And often, it requires both.

The question to ask before gathering more information

The next time you find yourself researching an important decision, try asking yourself something.

What would I need to learn that could actually change my decision?

It’s a deceptively useful question.

Suppose you’re considering an investment.

If learning that it could lose half its value would change your mind, then understanding its downside risk is important.

If learning that its fees are slightly higher than a competing product wouldn’t affect your choice, then spending another week comparing minor fee differences may not be worthwhile.

Suppose you’re thinking about building another source of income.

If your biggest concern is whether anyone would pay for your proposed solution, then researching website designs and business names probably isn’t the next useful step.

Finding out whether customers actually want the solution is.

The point isn’t to stop learning.

It’s to make learning serve a decision.

That requires three things.

First, knowing what you’re trying to achieve.

Second, understanding which information and assumptions actually affect your choice.

Third, identifying what you can do to reduce the remaining uncertainty to a level you can reasonably accept.

You may still decide to wait.

You may decide not to proceed.

Or you may discover that a small, reversible experiment is a better next step than a large, irreversible commitment.

All three can be perfectly sensible decisions.

The important thing is that you’re making them deliberately, rather than remaining trapped in an endless search for reassurance.

Perhaps we need a different measure of progress

For years, we’ve treated access to information as though it were almost synonymous with progress.

More books.

More courses.

More research.

More sophisticated tools.

And now, AI systems capable of producing answers to questions we might never previously have thought to ask.

These developments are genuinely valuable.

But the Endowus findings remind us that knowing something is important and having the information to pursue it aren’t the same as making progress toward it.

You can be well-informed about retirement and still have no workable retirement plan.

You can understand entrepreneurship and still have no evidence that your business idea will succeed.

You can know everything about the risks of changing careers and still be unable to decide whether to make the move.

Perhaps the more useful measure isn’t how much information we’ve accumulated.

It’s whether that information has improved our understanding sufficiently to make a better decision.

And whether we’ve acted on that decision.

Because information is only one ingredient.

Purpose tells us what we’re trying to achieve.

Understanding helps us interpret what we know.

Judgment helps us decide what matters.

And action allows us to discover what information alone cannot tell us.

The objective was never to know everything.

It was to know enough of the right things to make better decisions.

And perhaps the next time you feel stuck, the answer isn’t to learn something new.

It’s to look differently at what you already know.

For decades, entrepreneurship has been sold with a particular image.

Young. Technical. Plenty of energy. Few commitments. Willing to sleep on a couch, work all night and figure everything out along the way.

If you were 50, with a mortgage, a family and decades invested in a career, starting a business could feel like something you should have done 25 years ago.

But something interesting has happened.

The economics of starting a business have changed.

And some of those changes may actually favour the person who has spent the last 25 or 30 years working.

In March 2026 alone, at least four million people in the United States used ChatGPT to help plan, start, run or grow a business, according to an analysis by OpenAI.

And most weren’t building AI companies.

They were starting and running professional services firms, online stores, home and trade businesses, health and beauty practices, restaurants and other ordinary businesses.

In other words, AI wasn’t necessarily the business.

AI was, however, helping them build the business.

Because one of the perennial disadvantages of starting a business later in life was never a lack of useful knowledge.

It was the difficulty of turning that knowledge into a business.

What you have at 50 that you didn’t have at 25

At 25, you may have energy, time and an instinctive familiarity with new technology.

What you didn’t have is 25 years of accumulated experience.

You haven’t spent decades watching problems repeat themselves.

You haven’t seen good ideas fail because of poor execution, or mediocre ideas succeed because somebody understood the customer better.

You haven’t built the same professional network.

You probably had less capital.

And perhaps most importantly, you haven’t yet had enough time to discover how much of what looks obvious in a textbook isn’t obvious at all in the real world.

Those aren’t small advantages.

A recent AARP article on entrepreneurship after 50 highlighted precisely these assets: experience, financial stability and professional networks. It also pointed to free and inexpensive technology, including AI tools, as making entrepreneurship increasingly accessible.

There’s older research pointing in the same direction.

A study using U.S. Census Bureau data examined more than 2.7 million founders. The average founder was 42.

Among the fastest-growing one in 1,000 new businesses, the average founder was 45.

Even more interestingly, experience in the industry in which someone started a business was associated with substantially greater success.

So perhaps we’ve been asking the wrong question.

Perhaps the question isn’t:

Am I too old to start a business?

But instead:

What can I do today with everything I’ve spent the last 25 years learning?

Experience became more valuable at almost exactly the moment execution became cheaper

I think this is the shift many people over 50 haven’t fully appreciated yet.

Consider what used to be required to turn expertise into even a relatively simple business.

You might need someone to build a website.

Someone to design it.

Someone to write the copy.

Someone to create advertisements.

Someone to research the market.

Someone to help produce presentations and proposals.

Someone to handle routine customer communications.

Perhaps someone to help you understand spreadsheets, analyse data or work through unfamiliar software.

None of those things has disappeared.

And AI certainly doesn’t mean expertise no longer matters.

But the cost of getting from “I know something valuable” to “I can turn that knowledge into something another person can buy” has fallen dramatically.

OpenAI describes AI as increasingly functioning like a small business’s “first hire”.

It gives an entrepreneur access to capabilities that might previously have required employees, consultants or specialised software. Its data also shows entrepreneurs using it across very ordinary functions such as marketing, customer communications, sales and business development.

That alone changes the equation.

At 25, you might have had the advantage in execution. At 50, you had the advantage in experience.

Now technology is making execution cheaper. It isn’t making 25 years of experience cheaper.

AI doesn’t replace experience. It can leverage it.

This is where I think much of the conversation about AI gets things backwards.

We keep asking which jobs AI will replace.

For someone with decades of useful experience, it may be more productive to ask:

What can AI now enable me to do with what I already know?

AI can produce a marketing plan in seconds.

However, it doesn’t automatically know whether the plan makes commercial sense.

It can suggest 50 business ideas.

However, it doesn’t know which problem customers in your industry have been quietly complaining about for the last decade.

It can draft a proposal.

However, it hasn’t spent 20 years sitting across the table from customers.

It can generate options extraordinarily quickly.

But most importantly, options and judgment aren’t the same thing. And that becomes all the more important as the cost of generating things approaches zero.

When everyone can produce more ideas, more copy, more images, more analysis and more plans, simply producing things becomes less valuable.

Knowing which things are worth producing becomes more valuable.

That’s judgment.

And judgment is often the residue of experience.

There is another advantage to starting later

By 50, you’ve probably failed at a few things.

That’s useful too.

You know that enthusiasm doesn’t amount to a hill of beans.

Or that being good at something doesn’t necessarily mean people will pay you for it.

You know that revenue matters.

But that cash flow matters even more.

You know that smart people can be confidently wrong.

And you might have even learned (the hard way like I did) that one shouldn’t make an enormous irreversible bet.

In your investments and business, go slow. Take small steps. Get evidence and increase your commitment when reality gives you a reason to.

That may actually be a better way to use AI than asking it to magically invent your next business.

Start with what you know.

Look for a problem you’ve seen repeatedly.

Use inexpensive tools to test whether somebody will pay to have that problem solved.

Get proof.

Then build.

Of course, being 50 doesn’t automatically make you a good entrepreneur

Experience, on the other hand, can become baggage.

Expertise can make us overconfident.

A long corporate career can teach habits that don’t transfer particularly well to a tiny business where nobody knows your title and there isn’t a department down the corridor to solve the problem.

Whether you’re 25 or 50, the road to entrepreneurship requires learning.

And AI doesn’t remove the need to learn.

It doesn’t remove risk.

It doesn’t find customers automatically.

It certainly doesn’t turn a bad idea into a good business.

What AI changes dramatically, is the cost of finding out if a business can be a good one. You can research faster. Prototype faster. Create faster. Test faster. Learn faster.

And you can do many of those things before risking large amounts of money or walking away from your existing income.

That’s particularly important at 50.

You don’t have to behave like a 25-year-old founder to become an entrepreneur.

In fact, you probably shouldn’t.

Fifty isn’t the new 25

That’s not the point.

At 50, you don’t have what you had at 25.

You have something else.

You have perhaps three decades of mistakes, relationships, observations, expertise and pattern recognition.

Until recently, turning all of that into an independent business could require another set of capabilities you didn’t possess. Technical skills, marketing skills, design skills, software skills or enough capital to hire people who did.

That barrier hasn’t disappeared, but it has become substantially lower.

Which creates an interesting possibility.

For years, we assumed technology gave younger entrepreneurs an advantage over older ones.

Perhaps the more important effect of this generation of technology will be the opposite:

It allows experience to travel further.

And if that’s true, 50 may not be too late to start a business.

It may be one of the first times in history when the assets you’ve spent decades accumulating meet tools capable of helping one person put them to work at relatively low cost.

Experience became more valuable at almost exactly the moment execution became cheaper.

That changes what is possible.

By the time you reach your 50s, experience is supposed to be one of your greatest assets.

You’ve spent 20 or 30 years learning things that can’t easily be taught in a classroom. You’ve made mistakes. Solved difficult problems. Managed people. Dealt with difficult people. Seen good ideas fail and bad ideas somehow succeed. You’ve learned which problems deserve immediate attention and which will probably disappear if everyone stops panicking for 24 hours.

More importantly, you’ve developed something that is difficult to put on a résumé:

Judgment.

So it seems reasonable to assume that all those years of accumulated experience should make you more valuable in the employment market.

Except an uncomfortable number of people in their 50s are discovering the opposite.

In 2026, Business Insider interviewed over 100 people in their 50s navigating unemployment and career changes. Again and again, the same themes emerged: age discrimination, experience carrying less weight than expected, anxiety about AI and automated hiring systems, and employers simply ghosting applicants.

One of the biggest surprises was particularly revealing. People who had spent decades building careers were discovering that their experience didn’t seem to count for as much as they thought it would.

Perhaps there is another way to look at that.

What if your experience hasn’t become less valuable?

What if you’re simply trying to sell it to the wrong buyer?

The strange economics of experience

Think about what happens during the first half of a career.

At 25, you don’t have much experience. So an employer isn’t really buying what you know. They’re buying potential. They give you a job, provide systems, training, colleagues and opportunities, and over time you become more useful.

At 35, you can usually do considerably more.

At 45, you may be managing teams, clients, budgets or complex projects. You know the industry. You know where things tend to go wrong. You can often look at a situation and recognise something a younger colleague simply hasn’t encountered before.

By 50, you may have accumulated an extraordinary amount of knowledge, experience and judgment.

And yet something strange can happen.

The very experience that makes you capable of solving more valuable problems can make you harder to fit into the increasingly narrow box represented by a job description.

You may be too senior. Too expensive. Overqualified. Your experience may be in the wrong industry. Your previous title doesn’t quite match the title being advertised.

Or an automated system may decide that the words on your résumé don’t correspond closely enough to the words in the job description.

That doesn’t necessarily tell you what your experience is worth.

It tells you what that particular buyer is prepared to pay for it in that particular form.

Those are two very different things.

A résumé is a peculiar way to value 30 years of your life

Imagine trying to describe everything you have learned over the last three decades on two sheets of paper.

You list your positions. Your responsibilities. Perhaps some achievements. You try to quantify them wherever possible.

Then you send those two pages to an employer.

But they aren’t trying to work out everything you might be capable of doing with 30 years of accumulated experience.

They’re trying to answer a much narrower question:

“Are you the right fit for this particular role?”

That distinction matters.

The employment market is largely a matching market.

An organisation has already decided what it needs. It has defined a role, attached responsibilities to it and usually established a salary range.

Your task is to demonstrate that you fit inside that definition.

Which means much of the value you have accumulated may be irrelevant to the transaction.

Not because it has no value.

Because the buyer isn’t buying it.

There is another market for experience

Now imagine a completely different conversation.

Instead of asking:

“Who will hire me?”

Ask:

“What problems can I solve?”

Suddenly, many of the things that are awkward to communicate on a résumé become considerably more interesting.

Perhaps you spent 20 years in procurement and can immediately see why a small company is haemorrhaging money through its suppliers.

Perhaps you spent your career managing projects and can look at a chaotic business and identify exactly where things are getting stuck.

Perhaps you know an industry so well that you understand its customers, suppliers, terminology, economics and recurring problems almost instinctively.

Perhaps you’ve spent decades hiring people, selling, negotiating, managing operations, designing systems, dealing with regulators, organising events, teaching, writing, fixing machinery or doing one of thousands of other things.

Somewhere, somebody has the problem you learned how to solve.

That creates a very different economic relationship.

An employer asks:

“Are you the right person for this job?”

A customer asks:

“Can you solve this problem?”

Those are completely different questions.

And your age and ability to add value may differ greatly depending on which question is asked.

But there used to be another obstacle.

Even if you could identify a problem your experience allowed you to solve, turning that knowledge into something you could offer for sale wasn’t necessarily easy.

You needed more than expertise. You needed the ability to build websites, create content, design graphics, manage software, market yourself and somehow stitch all the technology together.

For someone who had spent 20 or 30 years building expertise in an entirely different field, that could be a significant barrier.

AI is changing that.

AI may be shifting the advantage

There is an understandable anxiety about AI among older workers.

Business Insider found that some people in their 50s worry both about AI filtering their applications before a human sees them and about being disadvantaged because they haven’t kept up with AI tools.

But that looks at AI only from the employment side of the equation.

Outside the employment market, AI may be doing something very different.

It is making many of the technical skills that once stood between your expertise and a potential customer considerably easier to access.

AI can help with research, writing, design, coding, analysis, administration and dozens of other tasks that previously required technical skills or outside help.

But AI still needs direction.

It needs someone to decide what problem is worth solving. What matters. What doesn’t. Whether the answer makes sense. Whether something will work in the real world.

In other words, as the technical barriers fall, judgment may become more important, not less.

And judgment is precisely what decades of experience can produce.

Your experience probably isn’t the product

Finding another market for your experience doesn’t mean trying to sell your experience itself.

Nobody wakes up in the morning wanting to buy your 27 years in procurement, operations, marketing or management.

They wake up with problems.

Your experience matters because somewhere inside those 27 years may be the ability to solve one of them.

And there are now more ways than ever to turn that solution into something another person can buy.

It might be consulting. A service. Training. A digital product. A small business. Or something you haven’t thought of yet.

The label isn’t particularly important.

What matters is identifying a problem you understand unusually well and determining whether there are people willing to pay to have it solved.

That is a much smaller question.

And, importantly, it is a question that can be tested before you make a dramatic career decision.

That doesn’t mean you need to become an entrepreneur

None of this means everyone in their 50s should abandon employment and start a business.

For many people, another job will be exactly the right next step.

The point is simply that employment isn’t the only market for what you know.

For most of your working life, you may have bundled your experience together and sold it to one customer. Your employer.

Your employer then combined your expertise with its brand, systems, technology, capital and other employees and sold the resulting value to its customers.

There is nothing wrong with that arrangement.

But it isn’t the only arrangement available.

Today, it is increasingly possible to take a small piece of what you know, attach it to a specific problem and offer the solution directly to the people who need it.

You don’t need to decide today that you’re going to “become an entrepreneur.”

However, you should ask yourself,

“What problem have I become unusually good at solving?”

That gives you a very different way to take inventory of your career.

Perhaps we’ve been measuring the wrong thing

There is something particularly brutal about being rejected for jobs after spending decades becoming good at what you do.

The Business Insider interviews capture some of that emotional toll. Respondents described frustration, damaged self-worth and the shock of encountering a hiring environment very different from the one they remembered.

It would be easy to draw a painful conclusion from that.

Maybe I’m simply worth less now.

But employment demand and economic value aren’t identical.

If nobody wants to buy your house for $1 million, that tells you something about the market for that house at that price.

It doesn’t mean the timber, land, bricks and everything inside suddenly ceased to exist.

Likewise, if employers aren’t responding to your résumé, it may tell you something important about today’s employment market.

It doesn’t erase what you spent decades learning.

You may need to start asking yourself,

Where is that experience most valuable now?

Because perhaps the great opportunity of your 50s isn’t trying to persuade someone that your previous 30 years still matter.

Perhaps it’s finally looking at those 30 years differently.

Not as a history of jobs you’ve held.

But as an accumulation of problems you’ve learned to solve.

And somewhere outside the confines of a job description may be someone for whom one of those problems is expensive enough to solve.

Your experience didn’t suddenly become worthless at 50.

You may simply need to find a different buyer.