[ad_1]
Psychologists have long known that the human brain succumbs easily to two forces. The ‘mere exposure effect’ shows that repeated exposure to the same idea makes us more likely to like and accept it. ‘Confirmation bias’ refers to our tendency to seek out and trust information that already matches our beliefs. These quirks of the mind explain why rumors persist and opinions become rigid.
But inside the architecture of social media, powered by AI-powered algorithms, they become supercharged. Platforms not only reflect our psychology – they weaponize it. Every like, every like, every share is not only training the algorithm; This in turn is training us, limiting what we see, intensifying what we are feeling and convincing us that those feelings are reality.
This cycle is no longer abstract. From the streets of Nepal to India’s communal riots and even the US Capitol riots, the mix of human bias and machine learning has become a feedback loop with political consequences. What began as a quirk of perception has been transformed into a system of dogmas.
The recent unrest in Nepal is a vivid reminder. What started as scattered online complaints quickly turned into protests. Social media platforms powered by recommendation engines captured and amplified user outrage. Within days, feeds became saturated with similar narratives, hashtags multiplied, and friend-recommendation systems coalesced angry voices into digital mobs. Then those crowds spilled out onto the actual streets.
It was not organized by any one group. It was organized by algorithms optimized for engagement. Each extra second of attention taught the AI what to promote next. In turn, that food trains humans to be angrier, more determined, more radical.
India too has seen how quickly a viral clip can harden opinions. From communal rumors on WhatsApp to polarizing videos on Instagram, digital content spreads across close-knit communities at lightning speed. When your entire digital circle is given the same clip, it doesn’t feel like “one perspective.” It feels like reality.
The upside here is clear: we think we are training the AI with our preferences. The truth is that AI is training us, telling us who to trust, who to distrust and even when to act.
The US Capitol riot of January 6, 2021 is another case study. Extremists may have sought out radical content, but recommendation engines intensified their visibility and increased their sense of collective power. In Europe, far-right and anti-immigrant networks developed in almost the same way, driven and clustered by algorithms. Large-scale studies complicate the picture: They show that algorithms do not radicalize average users. But for already vulnerable people, these systems are a speed-booster. And once groups come together, beliefs intensify. In short: People can light the spark, but AI stokes the fire.
Traditionally, “training AI” involves feeding it labeled data and refining it through feedback loops. Today, recommendation engines use the same model as humans – reinforcing content you engage with by showing it more often, gradually weeding out alternative viewpoints, and building communities conditioned around a narrow set of views that then feel universal. This is essentially the same logic of reinforcement learning, only this time turned away from its creators.
With over 800 million internet users and a demographically skewed youth, India perhaps faces the greatest risk. A nation where politics and religion are highly emotional cannot afford to let AI systems built in Silicon Valley shape its public imagination unchecked. Already, the IT-BPM sector makes up 7.4% of GDP, much of it dependent on US contracts and platforms. But the bigger risk is social: What happens when billions of micro decisions in Palo Alto start influencing beliefs in Patna or Pune?
If Nepal serves as a warning, India should act on the lessons by proactively building safeguards into its digital ecosystem. This means designing diversity into the feed to ensure the delivery of reliable, cross-cutting perspectives; adding friction to the virality to slow the wildfire spread of sensitive content; Auditing algorithmic clustering to prevent the creation of ideological silos; And educating users that each click not only teaches AI but also reshapes what it will teach tomorrow.
The greatest illusion of our time is that we are training machines. In fact, machines are training us to react faster, to trust more, to join the digital herd. Nepal’s unrest showed how quickly online conditioning can spill over into the streets. India and the world should consider this as more than a technical issue. This is a citizen.
Because when AI trains humans, the question is not what kind of technology we create, but what kind of society we become.
This article is written by Atul Rai, CEO of Stock Technologies.
[ad_2]
