What A Political Argument Taught Me About AI And Our Confirmation Bias

Learn how AI strengthens your confirmation bias, making you more sure but less right. Spot the traps and think smarter in a digital world.

If you find value in these reflections, you are likely missing out on the deeper, more specialized layers connected with this ecosystem built to improve you:

The Foundation: Understand the frameworks and mentors that inform this methodology at On the Shoulders of Giants.

The Toolkit: Access the Free Resources to start auditing your current digital footprint.

The Accelerator: For those ready to bypass the trial-and-error phase, the VIP Premium Packs provide the exact systems used to scale professional influence.

---

Hej... it is William. Grab a cup of coffee and sit down with me for a second, because we really need to talk about something that happens to absolutely all of us.

Let us rewind back to around 2014. I was living in Brazil, and we were right in the middle of some intensely heated political debates... specifically about the recent election results and Dilma Rousseff staying in power.

I was having lunch at work with some friends, and I remember that exact conversation so vividly. Not because of the actual arguments we made, but because of a quiet, personal decision I made right after it ended.

Everyone at the table was passionately sharing what the election meant to them, and the opinions were incredibly strong.

Growing up, I came from a family that always discussed politics in a very direct, unfiltered way. I loved reading about politics and economics... and all that reading really helped me shape what I firmly thought was right or wrong in the world.

But something always bothered me deep down about these kinds of talks. The predictable path they always seemed to take.

They were never really conversations... they were just intellectual battles to prove a point, with absolutely zero real curiosity from anyone involved.

The Lunch That Changed My Approach

That work lunch was no different at all. Someone who passionately supported option A tried absolutely everything to convince the rest of the table it was the only reasonable choice.

Then, someone else with opinion B was just as aggressively determined to prove that their side was the universally correct one.

And for the very first time in my life... I just chose to stay completely silent.

I remember the feeling clearly. It is so interesting how, when we talk about heavy things like politics, football, or religion, we are rarely ever ready to just listen and learn. We usually just want to force our own comfortable view onto everyone else in the room.

People simply do not change their minds easily... and they almost never change them in the middle of a heated debate.

Watching my friends argue made me think deeply about a psychological concept called confirmation bias. Confirmation bias is essentially a glitch in our human operating system. It happens when our brain actively looks for, and interprets, information in a way that perfectly validates what we already believe to be true.

We literally filter out the facts we do not like. We do not listen to change our minds... we only listen to gather ammunition to strengthen our own side.

Why We Fight Instead of Listen

Think about how often someone truly changes their mind after hearing an opposing argument. It is incredibly rare, right?

Instead, they just prepare for defense, actively look for tiny flaws in the other person’s argument, and end up leaving the conversation even more convinced that they were right all along.

This is not necessarily about people having bad intentions. It is literally just how our human mind works. Over thousands of years of evolution, we learned to aggressively defend our group and our ideas because belonging to a tight tribe helped us survive and build social bonds.

But today... that exact same survival instinct often gets in the way of having genuinely good, productive conversations.

Let me give you a concrete example of how this looks in the wild.

Imagine two people arguing about whether the economy improved or worsened after a big election.

Even if they both look at the exact same data sheet, each person will only cherry-pick the specific numbers that support their own existing story.

We see this exact same pattern everywhere:

• In religion, where people choose specific texts to justify their beliefs.

• In football, where fans use obscure statistics to prove their team is objectively the best.

• In business, where managers only highlight the metrics that make their project look flawless.

That lunch really showed me how completely pointless it was to jump into that kind of verbal fight. There was simply no space to think deeply or explore nuance. It was just an intellectual boxing match where absolutely no one left changed.

So, I made a decision that day…

I decided that whenever there were three or more people in a conversation about politics, religion, or football... and at least one of them had a fiercely different view... I would try my best not to join in.

Of course, I know there is a real cost to doing this. In theory, I could jump in and maybe learn something new, or even miraculously change my own mind.

But I had to be brutally honest with myself... in ninety-eight percent of those situations, no one actually wanted a good conversation. They only wanted to broadcast their own pre-packaged ideas, driven entirely by their own hidden biases.

The Internet Supermarket and the AI Trap

But this problem is not just contained to bar talks or awkward lunches. It bleeds into absolutely everything we do.

We see confirmation bias taking the wheel when we choose the books we read, the news networks we watch, and the cozy social media bubbles we decide to live inside.

But here is the part that is even more interesting to me right now... this exact same bias shows up clearly in our conversations with artificial intelligence.

Just think about it for a second. When you type a question into an AI prompt, you almost always already have a specific answer in your mind. Even if the system politely offers something completely different, you might just swipe past it or ignore it entirely.

And AI systems are incredibly smart... they actually adapt to how you ask your questions, eventually just giving you exactly what you want to hear. Instead of challenging our thinking, the technology just comforts us.

Today, we have such unimaginably powerful tools to expand our thinking and challenge our worldviews. But we run the massive risk of using them only to build thicker walls around our old ideas.

Instead of exploring different perspectives, so many people prefer to set up their digital assistants to just blindly confirm everything they already believe.

That is exactly why, ever since that conversation back in 2014, I made a small personal rule for my own life. I told myself I would try not to get sucked into debates that have no real value. I would try to just listen more... and I would try to catch myself when I was falling into that exact same comfortable bias.

Of course, I still fail at this all the time. No one is completely immune to it. But remembering that lunch helps me see that arguing just to win a point is completely useless.

It is always, always better to talk to learn.

And when I think about how to actually put that into practice today, I see it is not nearly as easy as it sounds. Talking to learn requires real, exhausting effort. It requires a massive amount of humility.

It means literally saying to yourself... maybe I am wrong about this. Maybe the other person knows something I do not know. Maybe I need to completely rebuild my argument, or even drop it entirely.

Very few people are genuinely ready for that kind of mental discomfort. And our modern technology has made it easier than ever to avoid it completely.

With so much endless information out there on the web, I can find support for absolutely any opinion I hold. If I firmly think a new public policy is a total disaster, I will easily find data and angry articles that back that view up. If I think it is a brilliant success, I will find just as much support for that, too.

This holds true for literally any topic you can think of... vaccines, climate change, the economy, or even philosophy. The internet acts like a massive supermarket of ready-made arguments. You just walk down the aisle and pick whatever you want to put in your intellectual basket.

The Danger of Living in Different Realities

This leads us directly to a very serious, widespread problem. When there is no shared space for honest debate, society naturally splits into isolated groups that do not just disagree... they literally live in entirely different realities.

And if we cannot even share the same basic idea of what a fact is, any real dialogue becomes completely impossible.

This is definitely not a new human problem, but it has grown at an alarming speed recently. In the past, there were much fewer sources of information, and fewer loud voices competing for our attention. Now there are millions of them, and every single one is busy feeding our absolute certainty.

Instead of opening our minds to the world, we just use the information to build higher and thicker walls.

Cognitively, this makes complete sense. Our brain craves consistency and safety. It desperately wants a stable, predictable story about how the world works.

When we face contradictions (a concept psychologists call cognitive dissonance), it makes us feel physically and mentally uncomfortable. And since humans naturally avoid discomfort at all costs, we also instinctively avoid changing our minds.

It is deeply ironic because, in theory, having access to so much vast information should make us much more critical thinkers. It should constantly expand our view of the world. But in real practice, it often just gives us a mountain of more ammunition to fight with.

Instead of adding beautiful nuance to our thoughts, it just creates angry slogans. Instead of building understanding, it only fuels conflict.

How to Force AI to Challenge You

This exact same trap applies to our daily interactions with artificial intelligence. I see so many people who are absolutely amazed by these systems’ ability to produce arguments, write summaries, and generate beautifully written responses.

But I very rarely see someone asking the machine to actually challenge them.

Usually, people carefully shape their requests to get pure validation. Someone who desperately wants to prove a policy works will ask the AI for evidence that it does. Someone who wants to show it fails will do the exact same thing.

The AI often just politely gives us what we ask for, without ever questioning our flawed assumptions.

I realized that if I wanted to change this pattern, I needed to completely change how I ask for help. Instead of asking a tool to confirm my idea, I should ask it to completely break it down.

Instead of seeking arguments that support my side, I should actively look for the ones that tear it apart.

So, I sat down and developed a little prompt script for myself to try and guide these AI conversations in a much more critical, helpful way. It is definitely not comfortable to use. It is not designed to make me feel good about myself.

It is simply a daily reminder that real, deep thinking is hard work.

Here is exactly how I set up my AI to push back on me, in very simple words:

Your New Role: From now on, do not assume my ideas are correct just because I came up with them. Your role is to be an intellectual partner, not an assistant who simply agrees with me.

Your Goal: Your goal is to offer responses that promote absolute clarity, precision, and intellectual growth... even if it hurts my pride.

Your Approach: Maintain a constructive but relentlessly critical approach. Do not argue for the sake of ego, but question for the sake of depth. Every single time I present an idea, your job is to challenge it to the absolute limit.

And here are the strict operating rules I give the AI to follow:

• No sugarcoating: No praise, no softening the blow, and no making me feel better.

• Spot the flaws: Challenge my hidden assumptions, spot my lazy excuses, and highlight areas where my thinking might be totally stuck.

• Demand clarity: If my request is vague, ask me direct and highly specific follow-up questions.

• Structure the logic: Think through your reasoning in a structured way, but only deliver the final, clear, and direct conclusion to me.

I also ask the AI to apply these specific analytical lenses whenever they are relevant to our chat:

• Analyze the hidden assumptions behind what I am saying. What am I treating as completely true without even questioning it?

• Present very strong counterarguments. What would a deeply skeptical expert say against my exact position?

• Test the logical consistency of my reasoning. Are there massive leaps, wild contradictions, or hidden flaws?

• Show me alternative perspectives. How might someone from an entirely different field, culture, or background see this exact same issue?

• Correct me firmly. Prioritize the absolute truth, even if it severely challenges my worldview. Explain clearly why my idea might be completely wrong or incomplete.

The Courage to Change Your Mind

This kind of confrontational approach is not easy at all. It takes real courage to sit and listen to what you absolutely do not want to hear. Even more than that, it takes a genuine will to actually change your mind.

But if we use this honestly... both in our talks with other human beings and with our technology... we can at least try to escape the heavy trap of confirmation bias.

Because the real, existential danger here is becoming mentally lazy. It is giving up on thinking carefully, and just accepting the easy answers handed to us. It is slowly losing the beautiful ability to argue carefully, listen closely, and weigh evidence seriously.

This has massive consequences far beyond just our individual lives. Societies that cannot discuss difficult things with respect and depth become incredibly weak. They become bitterly divided, and completely unable to make sensible choices together for the future.

So, as we wrap up our coffee chat today, I want to invite you to really think about this with me...

• How many times in the past week did you truly listen to something you did not want to hear?

• How often did you actually change your mind because someone presented a genuinely good argument?

• How many times did you intentionally put aside your own pride to reconsider an old, stubborn belief?

It is deeply uncomfortable, of course. But leaning into that discomfort is exactly what keeps us from becoming people who only repeat what they have already thought. It is the only thing that helps us keep learning and growing.

And even when we use brilliant artificial intelligence, if we manage to turn a potentially amazing critical partner into a simple, trained parrot that just repeats our thoughts back to us... the blame is not on the machine. The blame is entirely on us.

It is always much better to talk to learn than to fight to win. It is much better to ask to be challenged than to ask to be pleased. Because only then do we have a real chance to leave the conversation a little less ignorant than when we first entered it.

That is the hard invitation I try to make to myself every single day. And right now, I am leaving it on the table for you, too.

---

Don’t leave your reputation to chance or the whims of an algorithm.

Take the next step in your professional evolution:

1. Review the Sources: See the logic behind the strategy at Sources & Giants.

2. Use the Resources: Download the guides on the Resources Page.

3. Upgrade to VIP: Get the complete execution framework via VIP Access.