OPINION: Ravi Venkatesan
I have long been puzzled by something: How can smart, well-educated, decent people look at the same facts and come to completely different conclusions?
Someone who has read history and understands what happened in Nazi Germany or Rwanda can still cheer for a strongman who spews hatred. Someone who is kind to their neighbours can believe that a whole group of people doesn’t belong, or doesn’t deserve to be treated as fully human.
The easy, and wrong answer is that these people are stupid, ignorant or just bad. I have rarely found that to be true of the actual people, because they are often my friends, relatives, and colleagues at work whom I otherwise appreciate or even love. This has puzzled me for a long time.
Working with AI has given me a better way to think about this. An AI system, which, like us, is based on neural networks, learns from whatever it has been fed (trained on), and it then reads everything new through that lens (weights). We seem to work much the same way. Our parents, teachers, friends we hang out with, the books we read, our social media feeds, the communities we belong to, and the people we admire all teach us what to trust, what to doubt, who the good guys are, and which facts deserve our attention. Then new information arrives, and we don’t take it in fresh. We run it through the filters of what we already believe, which is why the same words or action can look like proof of genius to you and proof of tyranny and evil to another.
People who have studied the German Holocaust still argue about how ordinary people ended up taking part, and the arguments echo this. Hannah Arendt watched Eichmann’s trial and saw not a monster who sent a million people to their death but a diligent bureaucrat who never really stopped to think about what he was doing, or to see it through the victims’ eyes. She called it the banality of evil.
Daniel Goldhagen argued the opposite: that many Germans killed willingly because they had been raised on a deep hatred of Jews. Most historians think he blamed a whole culture too much. Christopher Browning, who studied a unit of middle-aged policemen ordered to shoot civilians, found something else again: men mostly going along with their comrades because standing apart felt too hard. Even among experts, there’s no single answer.
But the three stories describe three real ways we go wrong: We can be taught bad beliefs, we can stop thinking about our beliefs at all, and we can be swept along by the people around us.
However, there is a huge and critical difference between us and a machine. We can step back and look at where our own beliefs came from. That’s very hard, and most of us rarely do it.
I’ve had a small taste of it myself. For instance, for much of my life I believed in meritocracy. Give people a reasonable start, I thought, and the most talented and hardworking will rise, and that’s fair and as it should be. However, years of working in India’s districts showed me how much depends on where you start: your family’s income, the value system that shapes your values and education, whom you know, where you live, whether you can get a loan, and so on. Talent is everywhere. Opportunity isn’t. So the idea of meritocracy is much more complicated.
The philosopher Michael Sandel goes further. He argues that the idea of meritocracy leaves winners thinking they deserve everything they have and losers thinking they deserve their fate. And this mix of pride and humiliation, along with the contempt educated elites often show for people without degrees or money, has helped fuel much of today’s angry politics.
So here’s a clear example of needing to revisit one’s beliefs. It is called critical thinking.
As we move into a post-truth society, critical thinking- a willingness to step back and re-examine what we “know” or “believe” becomes incredibly important. So these are the questions I now try to ask myself.
Where did my views come from, and who shaped them? What do I read and watch day after day, and what is my social-media feed deciding I should see? Who am I surrounded by, and what would it cost me to disagree with them? Which of my beliefs have I never actually tested? And when I change my mind, am I learning something, or just finding better arguments for what I already thought?
In an age of AI, the danger isn’t only that machines learn the wrong things. It’s that we get better at defending what we already believe.
About the Author: Ravi Venkatesan Founder – GAME | Founder & Chairman – Institute of Human Flourishing | Chairman – GEAPP | Author
