Artificial intelligence is not necessarily creating the least intelligent generation in history. The greater danger is that people may gradually stop questioning what they see, hear and read.
We now live in a world where a convincing photograph may never have been taken, a familiar voice may belong to a machine and a confident answer may contain information that is completely false. The problem is no longer simply finding information. It is deciding what information deserves to be trusted.
That is why critical thinking is becoming one of the most important skills of the modern age.
Seeing is no longer believing
For generations, people were taught that seeing something with their own eyes provided a strong form of proof. Artificial intelligence has weakened that assumption.
Deepfake technology can reproduce a person’s face, voice and mannerisms with alarming accuracy. In January 2024, a finance employee at a Hong Kong company received a request to transfer approximately US$25 million. The employee became suspicious and attempted to verify the request through a video call with senior executives. The call appeared legitimate, but everyone on the call had been created using deepfake technology, and the employee was persuaded to complete the transfer.
This demonstrates why modern security cannot depend entirely on appearances. Important financial or organisational decisions should require several independent checks, such as calling a known telephone number, confirming through another authorised employee and following another established approval process.
Critical thinking begins when we stop asking whether something looks real and start asking whether it can be independently verified.
Authority should not replace evidence
People often assume that a claim must be reliable because it is supported by wealthy investors, respected professionals or influential organisations.
The collapse of Theranos demonstrated the danger of that assumption. The company attracted hundreds of millions of dollars in investment and assembled an impressive board of prominent leaders. Yet many of those individuals lacked specialist knowledge in blood testing and medical technology.
The company’s reputation, charismatic founder and ambitious promises created a powerful halo around it. Investors became so impressed by the people involved that they failed to examine whether the underlying technology actually worked.
Authority can provide useful guidance, but it is not evidence.
Whenever an organisation makes a major claim, the most useful question may be a simple one.
What would need to be true for this claim to be real?
That question forces us to look beyond presentations, qualifications and confident personalities. It directs attention towards testing, documentation, independent reviews and measurable results.
Words can be technically true and still misleading
Modern advertising rarely needs to tell a direct lie. Instead, it uses carefully selected language that creates a stronger impression than the available evidence supports.
Common phrases include “up to,” “starting from,” “as low as,” “clinically proven” and “recommended by experts.” These expressions may be technically accurate, but they often describe the most favourable possible conditions rather than the experience of an ordinary customer.
A product offering “up to 36 hours” of battery life may achieve that figure only under controlled conditions. A service advertised as “starting from” a low price may cost considerably more once essential features are included.
Critical thinkers do not automatically reject these statements. They place a question mark after them.
Up to 36 hours under what conditions?
Starting from that price with which features included?
Recommended by which experts?
These questions expose the distance between the marketing promise and the product being offered.
Agreement does not guarantee truth
People naturally look to others when deciding what to believe. This can be useful, particularly when a group contains people with relevant knowledge and independent perspectives.
However, it can also create groupthink.
Classic conformity experiments found that people sometimes repeated an obviously incorrect answer simply because everyone else in the room gave that answer confidently. Many participants later admitted that they knew the group was wrong but did not want to stand alone.
The same pressure now operates through social media. A claim may appear credible because it has thousands of likes, shares or supportive comments. But popularity measures attention, not accuracy.
When almost everyone agrees, it can be useful to deliberately search for the strongest opposing argument. This does not mean rejecting the majority or becoming suspicious of everything. It means understanding why reasonable people might disagree before reaching a conclusion.
AI can assist with this process when it is used as a challenger rather than an unquestioned authority. A person can ask an AI system to identify weaknesses in an argument, present the opposing case or list evidence that may contradict its first answer.
The final judgment must still belong to the human user.
AI should support thinking rather than replace it
Artificial intelligence can produce articles, reports, images, business plans and research summaries within seconds. This can save enormous amounts of time, but convenience has a cost when people accept the output without understanding it.
In an MIT study, participants who relied on ChatGPT to write essays showed lower levels of brain activity and struggled to recall lines from the work they had just produced.
Users should review the reasoning, check important claims against reliable sources and rewrite the information in their own words. They should also understand that asking several AI systems the same question does not automatically produce independent verification. Different systems may repeat the same errors because they were trained on similar information.
AI is most useful when it improves human thinking. It becomes dangerous when it quietly replaces it.
The hardest bias to challenge is our own
Not every distortion comes from technology, advertising or social pressure. Some come from within us.
People often accept information that supports what they already want to believe. We may ignore warning signs because we want a business opportunity to succeed, a political claim to be correct or a personal relationship to work.
This is known as confirmation bias, but its effect is often deeply personal. We do not deliberately deceive ourselves. We simply stop examining the evidence because the preferred conclusion feels more comfortable.
One of the most difficult questions a person can ask is:
What am I refusing to see because I need this to be true?
That question requires honesty rather than intelligence. It asks us to examine our motives alongside the available evidence.
Better technology requires better judgment
The age of AI will not be defined only by what machines can produce. It will also be defined by the decisions people make after receiving that information.
Critical thinking does not mean doubting everything. It means slowing down when the consequences are serious, examining the evidence, questioning persuasive language and recognising when emotion or social pressure may be influencing our judgment.
The world may continue to become more difficult to verify. Images will become more convincing, voices more realistic and automated answers more confident.
The solution is not to withdraw from technology. It is to become more disciplined in how we use it.
When the boundary between truth and fabrication becomes less visible, good judgment is no longer merely an academic skill. It becomes a form of protection.
This article is based on ideas and examples presented by Sandeep Swadia in his discussion on the crisis of critical thinking in the age of artificial intelligence.
