Google is taking a big risk by introducing AI into search results

Google will give you the answer to everything, even if it’s nonsense

Google’s integration of new AI features immediately caused a variety of problems with incorrect answers, from the absurd to the potentially dangerous. In short, what Google doesn’t know, AI will figure out.

Most AI features in mainstream products have so far been treated as optional extras. However, by placing AI-generated answers rank first in search resultsGoogle is signaling its intentions, even at the expense of the integrity of its most important product.

Google now responds on its own instead of providing a link to the publication

Previously, Google search results showed links to information, directing users to pages and summaries of other people’s content. Now, Generative AI in search not only presents information, but also summarizes and potentially edits it. The effects of these loose interpretations of reality can be misleading or even dangerous for the recipient. After the introduction of AI Overviews (available in the US for a few weeks now), Google was forced to make changes to improve results and explain the sources of blatantly erroneous summaries.

As it turned out, Google’s AI was able to feed off individual comments from Reddit and took them for granted. You could just as well treat the statements on Facebook groups as some kind of revealed truth. Meanwhile, we are not talking about some small project here, but about a search engine that is visited over 80 billion times a year and has over 90% market share.

Although Gemini and ChatGPT are steadily growing in popularity, their impact on humanity is much smaller than Google’s. Any stumble related to Google’s AI hallucinations risks the rapid spread of information and the loss of credibility of the search engine giant. The American company’s decisions may harm not only itself, but also globally harm solutions based on generative AI.