From information to insight: why knowing more does not automatically make you a better investor

The greatest change created by artificial intelligence may not be that machines are becoming more intelligent.

It may be that information is losing its price.

Annual reports, market data, summaries, comparisons and scenarios are available within seconds. Research that once required hours now appears on a screen on demand.

That sounds like an unlimited advantage. Until almost everyone possesses roughly the same information.

When access to knowledge no longer creates distinction, value shifts to the quality of your judgement.

Why this matters to you

Many investors try to reduce uncertainty by gathering more information. That can be useful. But an abundance of data can create three new problems:

the volume makes it harder to identify what genuinely matters;
a well-phrased answer is mistaken for a reliable conclusion;
almost any conviction can be confirmed more quickly.

AI accelerates this process. It reduces the cost of research, but not automatically the cost of a poor decision.

For a new investor, information becomes more accessible. For an experienced investor, old informational advantages disappear more quickly. For substantial wealth, governance, selection and context become more important than one more report.

The question therefore changes from how do I obtain more information?

It becomes: which information deserves to influence my decision?

Everyone receives the same ingredients

Imagine that every home cook receives exactly the same ingredients, quantities and recipes. One produces an excellent dinner. Another serves burnt pasta.

The difference is not the availability of ingredients. It lies in timing, experience, tasting, context and knowing when to depart from the recipe.

Investing works in much the same way.

AI can compare balance sheets, summarise quarterly figures and list potential risks. A sound decision must then ask:

Which information is already reflected in the price?
Which assumption determines the outcome?
Which source may be incorrect or outdated?
Which outcome does almost everyone expect?
What happens when many investors reach the same conclusion?

The information can be correct while the decision produces the wrong result.

First-order and second-order thinking

First-order thinking asks: what does this fact mean?

Second-order thinking asks: how will other people interpret this fact, and is that response already reflected in the price?

Consider a company that reports excellent earnings. The first conclusion is positive: higher earnings support value. But if the market expected even better figures, the share price may still decrease.

The fact was good. The surprise was insufficient.

Financial markets respond not only to information. They respond to the difference between information and expectation.

AI can help collect those expectations. It does not remove the problem. When millions of investors use the same tools, the same conclusions may become consensus more quickly.

A navigation app gives every driver the fastest route. Once everyone chooses it, that is where the traffic jam develops.

Answers are not the same as judgement

An answer describes. Judgement weighs.

An answer may say that debt has increased. Judgement asks whether the increase fits cash flow, the investment phase, interest-rate sensitivity and the share price.

An answer may list ten risks. Judgement identifies the two that could genuinely break the structure.

An answer may optimise a portfolio using historical data. Judgement asks whether the historical relationships will remain useful in the next crisis.

Experience therefore remains valuable. Not because an experienced investor knows everything, but because experience can make it easier to recognise which information is not decisive.

The new confirmation machine

AI is patient, persuasive and willing to develop almost any argument. That is useful. It is also dangerous.

Ask why a share is attractive and you receive arguments in favour. Ask why the same share is vulnerable and you receive arguments against it. The quality of the answer therefore depends partly on the quality and direction of the question.

Confirmation bias now operates at full speed. You are not consciously seeking agreement; you are simply framing questions that unconsciously support your existing judgement.

Use AI not only as an answer machine, but as an opponent:

Formulate the strongest reason why your analysis may be wrong.
Ask which evidence you are giving too little weight.
Assess the same case under different economic scenarios.
Ask which conclusion depends too heavily on one source or assumption.
Divide facts, estimates and assumptions.

The value does not lie in a longer answer. It lies in a better conflict between explanations.

An information protocol for investors

Step 1 — Define the decision first

Do not begin with endless research. State which choice you need to make and which information could genuinely change it.

Step 2 — Divide fact, interpretation and expectation

A revenue figure is a fact. The quality of that revenue is an interpretation. Future growth is an expectation. Do not give these categories the same status.

Step 3 — Find the strongest counterevidence

Do not look only for confirmation. Search for information that could disprove your analysis.

Step 4 — Check the source

A persuasive answer may rely on incomplete, inaccurate or outdated data. Verify important figures in primary sources.

Step 5 — Decide when information is sufficient

Additional research has diminishing value. Stop when new information is unlikely to change the decision—or when waiting becomes more expensive than the remaining uncertainty.

Step 6 — Record the judgement

Write down the assumption, the counterevidence, the remaining uncertainty and the reason for the position size. You can then assess later whether you genuinely learned anything.

The human advantage

AI does not feel loss aversion, but its user does. AI does not experience FOMO, but it can produce a persuasive story that justifies it. AI does not fall in love with a share, but it can help defend that affection intellectually.

The human advantage therefore does not lie in remembering more facts. It lies in:

understanding objectives and consequences;
weighing moral and personal context;
allowing doubt;
ignoring irrelevant information;
recognising behaviour before it becomes a decision;
accepting responsibility for the outcome.

These are not soft skills. They are elements of risk management.

The essence

AI makes information cheaper, but it does not automatically make sound judgement easier.
Markets respond to expectations and other people’s responses, not only to facts.
Divide facts, interpretations and expectations before drawing conclusions.
Use AI actively to find counterevidence, check sources and explore alternative scenarios.
Stop collecting when additional information no longer changes the decision.

In a world full of answers, the ability to ask the right question becomes scarcer. Perhaps the most valuable investor will not be the person who knows the most, but the person who knows what to ignore.

Further in Architecture of Wealth

The forthcoming e-book Architecture of Wealth examines how information, behaviour and decision-making can form a repeatable process. Not to replace human judgement, but to make it more open to scrutiny.