Unlimited curiosity, uncertain outcomes: The risks of AI-augmented mass inquiry

Human curiosity is abundant, but the means to pursue it remains scarce for most. A bright farmer, a self-taught learner in a poor town, or a first-generation student shut out of elite institutions could ask profound questions yet lack tools, access, time, or networks to act on them. That gap shaped civilisation.

Institutions, markets, and political systems evolved around a simple constraint: only a limited number of curious minds could turn inquiry into consequential action at any given time. Progress moved, but it moved with a governor.

Artificial intelligence is beginning to weaken that governor. The implications are neither simple nor settled. They may be liberating, destabilising, or both. We are entering terrain with few historical maps.

More from Opinion

The first disruption is cognitive and social. Curiosity often begins with the awareness of what we do not yet know. AI can shorten that distance with startling speed. It can explain difficult concepts, summarise research, compare legal arguments, translate languages, generate code, and tutor users at low cost. Capabilities once reserved for specialists are moving into ordinary hands.

But wider access should not be confused with the end of gatekeeping. Older systems relied on universities, publishers, professional bodies, and credentialed experts. Newer systems rely on models, platforms, ranking systems, moderation rules, and a small number of companies that control the computing stack. Authority does not vanish. It changes costume.

History suggests caution. The printing press widened literacy and also widened conflict before it nourished the Enlightenment. The internet democratised speech and democratised misinformation with equal enthusiasm. Major technologies often empower citizens and unsettle institutions in the same motion. AI is unlikely to be the first exception.

The second disruption is physical. AI-powered curiosity is not weightless. Mass curiosity, when AI-mediated, is not a purely psychological event. It is a resource extraction event. Every query runs through data centres, semiconductors, electricity grids, cooling systems, and mineral supply chains. Each polished answer on a screen rests on industrial machinery most users never see. If billions of people begin using advanced AI routinely, demand for computing power will rise sharply.

That demand carries costs. Chip manufacturing depends on materials such as cobalt, gallium, germanium, and rare inputs concentrated in geopolitically sensitive regions. Data centres require heavy electricity loads and significant water for cooling. In water-stressed regions, that is not a technical footnote. It is a public question. The benefits of AI may be global, but many of the burdens will remain local. Some populations promised digital inclusion may also carry a disproportionate share of the environmental and extractive costs.

The third disruption is institutional. Courts, regulators, universities, peer review systems, and democratic processes were built for a world where informed challengers emerged in manageable numbers. They were designed to absorb criticism, reform, and contestation at a human pace. What happens when millions of people can use AI to interrogate contracts, challenge expert claims, test public policy, audit bureaucracies, or expose inconsistencies at the same time?

Many institutions deserve scrutiny. Some have become complacent, exclusionary, or slow. Yet reform and destabilisation are not synonyms. A flawed institution can still collapse faster than it can improve. That risk deserves more attention than it receives.

Then comes the psychological tension. Curiosity, once actionable, does not stay neutral for long. It acquires direction: political, ideological, tribal, personal. AI can sharpen motivated reasoning by giving elegant answers to loaded questions. It can flatter confidence without building understanding. That could produce a civilisational version of the Dunning-Kruger effect: more people certain of partial knowledge, more camps armed with internally coherent but mutually hostile belief systems.

Yet that is only half the story. AI can also become the most patient tutor many people have ever had. It can challenge weak reasoning, explain trade-offs, surface contrary evidence, and insist on nuance when human instinct prefers slogans. The real contest may not be between humans and machines. It may be between two uses of the same machine: one that deepens tribal certainty and another that raises public reasoning.

No clean conclusion follows. Better chips, cheaper inference, smaller models, edge computing, and cleaner energy may reduce some of these pressures over time. Institutions may adapt faster than sceptics expect. Citizens may use these tools more responsibly than critics fear.

Still, one fact is already visible. The modern world was built for an uneven distribution of capability. Expertise was concentrated. Access was scarce. Curiosity often stalled before it became consequential. That arrangement is changing faster than the surrounding systems can comfortably absorb.

If large numbers of ordinary people gain the power to analyse, create, challenge, and organise at much higher levels, societies will need stronger institutions, wiser governance, and broader civic maturity. Technology alone will not supply those things.

Whether this era becomes a renaissance, a period of disorder, or the familiar mixture of both that follows every information revolution will depend less on AI itself than on decisions being made now, often by people who have not yet grasped the scale of the shift.

That may be the most consequential uncertainty of all.

(Sreejith Sreedharan is a technology analyst and author. The views expressed in this piece are personal and solely those of the author. They do not necessarily reflect Firstpost’s views.)