AI can answer questions in seconds. It can summarise 50 pages, generate arguments, challenge an idea, write code, compare alternatives and even critique its own response.

<aside> 💡

So, if AI can do all of this, do humans still need critical thinking?

</aside>

I think the answer is the opposite of what we might expect. Critical thinking is not the same as intelligence. We often imagine critical thinkers as people who are exceptionally smart or knowledgeable.

But researchers generally describe critical thinking differently. It involves skills such as analysing evidence, evaluating claims, drawing reasonable conclusions and, perhaps most importantly, monitoring and correcting your own thinking.

That last part matters.

Because one of the biggest obstacles to good thinking is not lack of intelligence. It is our tendency to defend conclusions we already like. Research on confirmation bias and “myside bias” has repeatedly shown that people evaluate information differently depending on whether it agrees with their existing beliefs. And higher cognitive ability does not automatically make this tendency disappear. In other words:

<aside> 💡

Being good at reasoning does not necessarily mean being good at reasoning against yourself.

</aside>

Can we actually improve critical thinking? Fortunately, yes. But probably not by memorising a list of logical fallacies.

A major meta-analysis by Philip Abrami and colleagues examined 341 experimental and quasi-experimental effects on critical-thinking instruction. Overall, teaching critical thinking produced a measurable improvement, with an average effect of about 0.30 standard deviations. (Sage Journals)

More interestingly, some of the most useful approaches involved:

• discussing and defending ideas through dialogue
• solving realistic, authentic problems
• receiving feedback and mentoring
• explicitly reflecting on how we reached a conclusion

How can we become better critical thinkers?

For me, the research translates into these five habits.

1. Look for evidence that could prove you wrong.

   This turns reasoning from defence into investigation.

2. Separate confidence from evidence.

   Feeling certain is not the same as having strong evidence.

3. Make your reasoning explicit.

   Before making an important decision, write down:

   • What do I know?
   • What am I assuming?
   • What alternatives have I considered?
   • What evidence supports each explanation?
   • What am I missing?

4. Build domain knowledge.

   Critical thinking is not completely independent of knowledge. You cannot critically evaluate a cybersecurity architecture without understanding cybersecurity, or a clinical study without knowing something about experimental design. General reasoning principles help, but they become powerful when combined with deep knowledge.