In the previous chapter we saw how to research with AI without falling for its traps. Now we take the most valuable step and, curiously, the least used: turning AI into a counterweight to your own thinking. Because the best investment an AI can make in your finances isn’t telling you what to buy, but helping you not fool yourself. And humans, with money, fool themselves constantly.

Honest comparisons between options

Let’s start with the most direct. When you’re torn between two or three options — two funds, rent or buy, overpay the mortgage or invest — AI is excellent at structuring the comparison, as long as you provide the data.

A useful prompt: “Here’s the data for these options [data]. Compare them in a table by cost, risk, liquidity, horizon and effort. Don’t pick one for me: show me where each wins and what kind of person would prefer each option.”

The key is that “don’t pick for me.” What you want isn’t a verdict, but to see the full map of pros and cons laid out, without one detail blinding you to the rest. AI is great at stopping you falling in love with a single number (the return) and ignoring the others (risk, liquidity, fees). For the fine numerical work, combine it with calculation tools like our Compound Interest Simulator.

Simulate scenarios: the “what if…?”

Here AI does something the human mind struggles with: thinking coldly about what could go wrong. Before an important decision, put scenarios to it:

“If I make this decision [decision], simulate three scenarios: a favourable one, a normal one and an adverse one. In the adverse one, what would happen if the market drops 30%, if I lose my job, or if I need the money sooner than planned? How serious would each be?”

This exercise is gold because it attacks the default optimism with which we make financial decisions. We tend to imagine only the good scenario. AI forces you to look at the bad one — not to paralyse you, but to check whether you’d withstand it. An investment that looks great while everything rises may be unbearable for you in a fall; better to discover it by simulating than by living it.

Ask especially about the scenario that hurts most: “what would have to happen for this to be a bad decision?” If the answer is something plausible and you couldn’t bear it, you’ve just avoided a mistake.

AI as devil’s advocate

In the previous chapter we warned about the “complacent yes”: AI tends to agree with you. The good news is you can flip it and turn that flaw into your greatest ally, by explicitly asking it to push back.

“I’m going to make this decision [decision] and I’m fairly convinced. Play devil’s advocate: give me the three strongest arguments against, the risks I’m underestimating, and what a prudent person who doesn’t share my enthusiasm would tell me.”

Actively asking to be challenged is one of the most valuable habits there is, and almost nobody practises it because it goes against the ego. Having an infinitely patient tool willing to play critic, without it stinging the way a person would, is a luxury. Use it especially when you’re very sure: excessive certainty is precisely when judgment fails most.

Watch your biases, don’t amplify them

With money, our brain has systematic traps. AI, used well, helps you spot them; used badly, it reinforces them. It’s worth knowing the main ones:

  • Confirmation bias: you seek data that proves you right. If you only ask AI to support your idea, you fall straight in; that’s why the devil’s-advocate exercise matters so much.
  • Recency bias: you believe what just happened will keep happening (if it rose, it’ll keep rising). Ask AI for historical perspective to counter it.
  • Loss aversion: losing hurts more than gaining feels good, leading to panic-selling or to avoiding reasonable risk.
  • FOMO: the fear of missing out, especially with investment fads. AI can help cool the euphoria by calmly explaining the real risk.

A powerful prompt: “I’m going to make this decision [decision]. What cognitive biases might be influencing me, and how would a rational person correct for them?” Having a tool name your own biases, without judging you, is a surprisingly effective way to recover financial sanity.

The decision is still yours

This whole chapter points to one idea: AI is more valuable as a critical mirror than as an oracle. It compares for you, simulates for you, pushes back and flags your biases. But notice it never decides.

And so it should be, because the decision integrates things AI doesn’t have: your real risk tolerance, your goals, what keeps you up at night, your full situation. AI helps you reach that decision more clear-eyed and less deceived, which is a great deal. The last step — pressing the button — is yours, and it’s where your responsibility as an investor lies.

With that we close the decisions block. In the final block we move from deciding to operating: how to automate repetitive financial tasks with AI to save time, and how to do it without risking the most valuable thing you hand a machine — your data.