Google's AI Overviews search feature mistakenly identified the year as 2024, leading to mockery on social media. This incident emphasizes the ongoing problem of AI hallucinations, where language models occasionally provide incorrect information. Users recalled previous outrageous errors from the feature, leading to skepticism about its reliability. In response, Google stated that they are actively working to update their AI systems to reduce such mistakes, though the company did not clarify the cause of this misunderstanding.
Simply asking AI Overviews if it's 2025 resulted in a head-scratching answer. "No, it is not 2025," the AI responded, leading to widespread mockery on social media.
The humiliating instance highlights a persistent and glaring problem plaguing some of the most advanced large language models: frequent hallucinations remain an enormous - and growing - issue.
Rocks and Glue Google's AI Overviews, in particular, has garnered a reputation for consistently leading users astray. The feature made headlines last year for outrageous missteps.
As with all Search features, we rigorously make improvements and use examples like this to update our systems," a spokesperson told TechCrunch.
Collection
[
|
...
]