AI Search and the Atrophy of Human Intellect

It’s fascinating how ‘Six Degrees of Separation’ works. 

For the uninitiated, it’s a theory that dictates that any two people in the world can be connected through a chain of six other folks, at most.  Technically speaking, you could find a phantom thread between Paul Thomas Anderson and yourself. 

Ideas work the same way. Any concept you’re mildly curious about can be the beginning of an accidental education of sorts. That Wikipedia article on sleep paralysis may be four links away from some medieval demon folklore. That recipe for Thai curry? It could be five clicks from the history of colonialism and spice routes. Nearly everything connects at some point if you’re willing to follow the thread. The internet is a goldmine of such rabbit holes. But it may not be for much longer. AI search is here, and you may soon have to say goodbye to the thrill and joy of discovery.

The Numbers Tell Us a Sobering Story

Chatbots like ChatGPT, Gemini, and Claude are rapidly making organic search redundant. They deliver answers to complex questions with a single prompt. This removes the incentive to visit a search engine and discover information organically, particularly from primary sources. 

Even for users who manage to circumvent the trap of AI search, Google’s AI Overview feature gets in the way. It hijacks the results and directly offers answers in a neat little snippet. 

So let’s take stock of the damage:

• With AI summaries thrown into the mix, users clicked on traditional search results only 8% of the time. Without these summaries, the preference for organic search nearly doubled to 15%.

• News-related searches on ChatGPT shot up by 212% between January 2024 and May 2025, directly siphoning traffic from news portals. 

• Google’s AI Overview also ruthlessly cuts into organic traffic. Sites that enjoyed the top spot in search results could lose around 79% of their traffic when pushed below the overview segment. 

Google spokespersons, however, have denied these claims. In her blog post, Google VP and Head of Search Liz Reid writes: 

“…average click quality has increased and we’re actually sending slightly more quality clicks to websites than a year ago. […] This data is in contrast to third-party reports that inaccurately suggest dramatic declines in aggregate traffic — often based on flawed methodologies, isolated examples, or traffic changes that occurred prior to the roll out of AI features in Search.”

Whether this is a debate over data or a difference in interpretation, the impact on content discovery is devastating. Publishers are bleeding revenue. Organic traffic is collapsing across the board. And websites with painstakingly well-researched articles are struggling to attract audiences. 

But internet economics is not the only victim. AI search is also systematically dismantling the very way we search for information. 

The Death of Discovery

Organic search is not entirely flawless. It’s manipulated by SEO and dominated by big publishers. Still, it preserves intellectual autonomy, which is critical to the human intellect. 

When you type a question manually into an organic search engine, you have multiple sources at your disposal. This gives you an opportunity to: 

• Compare them and make quick choices
• Notice contradictions and biases
• Recognize credible publications 
• Skip the unreliable sources
• Become intimately familiar with the nuances of a subject or topic

It is in these gaps between searching and finding, where you’re forced to exercise judgment, that critical thinking is born and built. 

AI search, however, eliminates large parts of this process. Core steps like reading, doubting, and understanding are entirely snipped off. This is unfortunate because the human mind is shaped by the act of learning — the very thing we’re losing to AI search.

The Importance of Primary Sources

Outsourcing learning to AI search also makes us unconcerned with the quality of information we absorb. Primary sources, once considered the holy grail of journalism, are now increasingly being phased out in favor of AI-friendly pages. 

Sure, these tools can summarize content and capture the essence of primary sources. But how complete and accurate are these rehashed answers? They are, at best, effective summaries. And by their very nature, summaries involve loss of information. 

When we skip past a creator’s original commentary on their work and settle for an AI chatbot’s version of it, we lose access to the bias and brilliance behind their art. When we trade the chance to read a critic’s original review for a watered-down, AI-generated overview, we fail to learn the nuances behind their analysis. 

All we have is the conclusion without the reasoning. The verdict without the trial. 

Beyond the text itself, context matters enormously in the search for credible information. This context is built through numerous subtle cues scattered across digital spaces. Like a website’s visual design, its editorial standards, and the sources it cites.

These signals help us understand how much we can afford to trust what we’re reading. AI search strips all of this away. The information arrives decontextualized and homogenized in a familiar black box. We have no idea where it came from. Consequently, we lose the ability to sniff out quality. 

The Anatomy of a Vicious Feedback Loop

The economic implications of AI search can be catastrophic for quality journalism. Right now, publishers survive on traffic. People search, click through, read articles, and see ads that generate revenue. That’s the current model. 

But if an AI tool can summarize a high-quality, 3,000-word investigative report in a chat window, why would anybody visit the original site? No visits means no revenue. And no revenue means no investigative journalism. 

The only content that can survive in this ecosystem is content that doesn’t need to make money. Think corporate PR releases, propaganda from governments, and whatever low-effort slop that can be produced cheaply enough. The business model for serious information collapses, leaving us stranded in a wasteland of subsidized messaging. 

The nightmare doesn’t end here. When websites cannot monetize through traffic, they are forced to optimize for the only audience that matters — the AI itself. So, AI models train on stolen web content. New web content gets optimized for or generated by said AI models, which then train on synthetic AI content. 

This leads to a feedback loop with potentially terrifying implications. Each cycle feeds itself, limits the pool of information we have access to, and cuts off our access to any real knowledge. The information pool becomes shockingly limited and grows progressively worse till we collectively become dumb and dumber. 

This absurdity is best captured by riffing on Dr. Ian Malcolm: “The human brain creates AI. AI destroys the human brain. AI inherits the earth.”

Who Owns the Truth?

AI may inherit the earth, but as things stand now, it has no claim on the truth. Its results are shamelessly easy to manipulate if you hold the power to do so. 

Grok by xAI is a prime example of this inherent weakness. Elon Musk has been guilty of tweaking Grok’s training data to produce more right-leaning results. Now, whether you agree with that specific bias or not, the mechanism itself should be alarming, to say the least. 

Someone, somewhere, at every AI company, is making decisions about what sources are authoritative, what framing is neutral, and what information is reliable. These decisions are not objective. They occur in closed systems behind closed doors, invisible to the end user. We have no way to measure the extent of the bias involved. Left unchecked, this could allow AI to successfully transition from product to propaganda.

The whole thing feels strangely Orwellian. Today, a handful of private entities are responsible for collecting, consolidating, and commodifying all the information at our disposal. How much further before they start to distill all the information that goes out into the world? 

Reclaiming Our Intellect in an Increasingly Thought-Agnostic World

The proliferation of AI tools has made us unwilling to scroll past even the first page of organic search results. Now, with seemingly legitimate links provided to us in a chat window, it’s tempting to take the path of least resistance and give up the joy of thinking entirely. 

It doesn’t help matters that these tools are available at no cost. We may know, in theory, that if something is given to us for free, we are the product. We’ve interpreted this to mean that we — meaning our data and personal information — are the price we pay for free access to AI. 

Turns out, that’s not true at all. The real cost is our intellect. 

So, how do we go about retaining and reclaiming our intellectual autonomy in today’s world?

1. Cast a Wider Net

The obvious starting point is to remove the blinders and broaden our sources of information. Log out of your AI chat account for a week. Exclusively use organic search and skip the AI overview. Go beyond the first page of search results. And if it is at all feasible, consider subscribing to high-quality paid content online.

2. Go Analog

It helps to take a digital break and converse with people offline. Human dialogue is critical to our collective thinking. It is how we pass on our stories, and in our stories lie nuggets of wisdom that cannot be claimed by AI crawlers. 

3. Write

At least, attempt to. Write for an audience of one, if you must. Even if you are the only person who will read your work, make the effort to externalize your ideas. It is a definitive way to assess your proficiency in a subject and identify gaps in your knowledge. To examine if your opinion on an issue is really yours, or if you’re merely repeating something a bougie autocomplete tool told you.

---