Can We Trust AI-Powered Web Searches?

It's becoming increasingly apparent that Google and its rivals are adopting AI-generated summaries. However, research indicates that these results are far from infallible and can be manipulated. Should this be a cause for concern?

AI and Chatbots: A Biased Potential?

Does aspartame cause cancer? The potentially carcinogenic properties of this popular artificial sweetener have been debated for decades. Last year, the World Health Organization labeled aspartame as “possibly carcinogenic” to humans, while public health regulators claim it’s safe in the small amounts typically used. But in today’s digital age, is a simple Google search enough to clear up such controversial debates?

The Era of Chatbots

Tech companies have touted AI chatbots as a utopian tool capable of replacing various jobs and services – including search engines. Instead of scrolling through a list of web pages to find an answer, an AI chatbot can, in theory, browse the internet for you, compiling relevant information to fulfill your query.

The Question of Objectivity

The user-friendliness of this method has raised questions about the criteria chatbots use to select information. Researchers have found that these tools tend to overemphasize texts loaded with technical jargon or keywords, while overlooking other “trustworthy” elements, such as the presence of scientific references or objective language.

The Stakes of the Future

While these criteria may suffice for simple queries, they pose a problem for more complex debates. What if someone could “hack” these criteria, potentially swaying the information presented to billions of internet users?

As technology leaps forward, so do methods for “manipulating” chatbot results. This raises significant ethical questions that require deep reflection. Users need to be aware of the techniques at play and the implications they might have on the information they receive. For a reliable and unbiased internet, vigilance is essential.