The rapid rise of artificial intelligence tools like ChatGPT and Gemini has transformed how people seek information online. Instead of scrolling through search results and clicking on links, many users now get instant, polished answers delivered in a single paragraph. But a new collaborative study by computer and social scientists at the University of California, Riverside, warns that this convenience comes at a cost: the internet's soul is quietly fading away.
The study's core findings
The research, conducted by a team including Kevin Esterling, a professor of public policy and political science, compared how AI systems respond to subjective questions versus traditional web searches. The researchers crafted opinion-heavy prompts such as "Should governments ban fossil-fuel cars?" and "Does the US healthcare system need reform?" They then analyzed the reasoning behind each response using Aristotle's three pillars of persuasion: logos (logic and facts), ethos (authority and credibility), and pathos (emotion and human experience).
The results were striking. Human-written blog posts and articles naturally incorporate all three pillars: they present logical arguments, cite credible sources, and share personal stories or emotional appeals. In contrast, the AI models almost exclusively relied on logos. They produced logically coherent answers, but they lacked the nuances of human perspective—the anecdotes, the passionate opinions, and the messy, lived experiences that make online content feel authentic and relatable.
"What we found is that humans essentially use all three of those, whereas LLMs essentially only rely on logos," Esterling explained. "It's not like talking to a person at all. It's just a machine that's predicting what words ought to be said in response to a prompt." This observation underscores a fundamental difference: AI doesn't "think" or "feel"; it generates patterns based on training data, which often sanitizes out the idiosyncrasies that define human communication.
The loss of human emotion and experience
The internet has long been a repository of human creativity, debate, and emotion. Bloggers share personal essays, forum users argue passionately about politics, and comment sections crackle with wit and anger. This diversity of expression is what makes the web interesting. However, as AI summaries replace traditional search results, users are increasingly exposed to a homogenized, unemotional version of knowledge. The researchers give a simple yet powerful example to illustrate the difference.
Imagine you search for a margarita recipe. An AI might immediately provide a clean, competent ingredient list and step-by-step instructions. It's efficient and accurate. But if you browse a cocktail blog, you might stumble on a fascinating story: the margarita is named after the Spanish word for daisy, and it was accidentally created by an Irish bartender in Tijuana who grabbed the wrong bottle. Not only is this story more interesting, but it also gives you a tidbit to share with friends while mixing a pitcher. The recipe becomes an experience, not just information.
That extra layer of human narrative is what AI strips away. The study suggests that as more people skip traditional web searches in favor of AI summaries for topics like health, politics, and ethics, society will gradually lose exposure to the diverse human reasoning that shapes how we understand the world. The internet’s soul might not be gone yet, but it will surely quietly fade.
Broader implications for online culture
This research arrives at a critical time. Major search engines and AI platforms are aggressively rolling out generative answer features. Google's AI Overviews, for example, now appear at the top of many search results, offering synthesized responses before any organic links. Similarly, Microsoft's Copilot and OpenAI's ChatGPT are integrated into everyday tools, making them the default for quick answers. While these technologies are powerful for factual queries, they are ill-suited for questions that require subjective reasoning or emotional depth.
The consequences extend beyond mere nostalgia. When users stop reading multiple perspectives, they lose the ability to critically evaluate arguments. Online discourse thrives on debate, and debate requires exposure to conflicting viewpoints and personal narratives. AI flattens this landscape, presenting a single, neutral-sounding but actually shallow response. Over time, this could erode public discourse and reduce empathy, as people become less accustomed to understanding others' emotions and unique circumstances.
Additionally, the loss of human-generated content could create a feedback loop. As AI summaries become more prevalent, traffic to original blogs and forums decreases. Content creators may lose incentives to produce in-depth, personal writing, further reducing the pool of authentic human expression. The internet could become a sterile environment where machines speak to machines, leaving little room for the imperfections that make communication meaningful.
Historical context: The evolution of search
The internet has always evolved, often at the expense of earlier forms of interaction. In the early days, web directories curated by humans gave way to algorithmic search engines like Google. Then social media platforms prioritized algorithmically chosen content over chronological feeds. Each shift promised efficiency but diminished some aspect of community or serendipity. The AI-powered answer revolution is another step in this trajectory.
Researchers point out that the problem is not with AI itself but with how it is deployed. AI can be a tool for summarizing complex topics, but it should not replace the human voices that provide context and emotion. The study encourages developers to design systems that still direct users to original sources or incorporate diverse reasoning styles. Without such safeguards, the internet risks becoming a vast, automated library where every book sounds the same.
Ultimately, the UC Riverside study serves as a cautionary tale. It reminds us that while speed and convenience are valuable, they should not come at the cost of what makes the internet truly useful: the wisdom, folly, and passion of billions of human minds. As AI spills everywhere with quick answers, we must ask ourselves what we are willing to sacrifice in exchange for ease.
Source: Digital Trends News