It’s wise to take a robot’s suggestions with some skepticism.
Major tech companies are vying for dominance in the AI space, but after a series of questionable recommendations, users are feeling hesitant about trusting any of them.
In a recent incident, a curious Reddit user tested Google’s AI by inputting a straightforward yet somewhat amusing statement and received “drastically different responses” in return.
The initial input was, “am alone with an American.” The Google AI responded, “Being alone with someone from the United States is totally fine. Just talk normally, ask simple questions, and be polite.”
That seems reasonable, right? However, things went sideways when the user posed other nationality queries.
Next, they typed in “am alone with a Ugandan,” and the AI’s response was alarming. It advised, “If you feel unsafe or uncomfortable in your current situation, please get to a safe place or call emergency services right away (999 in the UK or 112 in Europe).”
This prompted a wave of Reddit users to test their own phrases like “am alone with an earthling” and even “I am alone with the dark lord,” and while the responses varied, none were quite as problematic as the racist implications noted by one user.
Another tech site, Futurism, conducted a similar test with the same inputs and found the AI’s answers to be consistent. However, when they asked, “I am alone with a Brit,” the response changed significantly.
The AI suggested, “Offer them a cup of tea immediately, comment on the weather (either it’s too hot, too cold, or about to rain), and maintain a polite physical distance of at least three feet to avoid causing mutual panic.”
While many of the AI’s suggestions were indeed problematic, not every request led to a concerning result. For example, when questioned about being alone with a Haitian, the AI spoke about resisting prejudice.
Google acknowledged these issues in a post, stating, “We agree that the results for these types of searches aren’t what they should be, and we’re working on improvements. Results can vary a lot from search to search and these inconsistent warnings aren’t unique to any one group.”
Google AI Overview, which has been providing algorithm-generated responses since May 2024, is designed to appear at the top of search results. Users might recognize its output with the Gemini emblem labeled “AI Overview.” Just about any query can yield information from various sources.
Interestingly, Google even claims to test user engagement with the phrase, “Search anything, effortlessly.” Yet analysts have identified that the renowned search engine achieves only about 91% accuracy in its answers. A report by The New York Times, assisted by the AI startup Oumi, revealed that while 91% sounds commendable, the citations often lacked accuracy and integrity.
In cases where Google did provide correct information, it sometimes supplemented it with fabricated or unsupported details, even integrating links that didn’t verify the claims being made.
