Google's AI Overviews feature has come under intense scrutiny following a series of bizarre and incorrect suggestions. The feature, which was recently announced in the US, has gone viral on social media for recommending actions such as adding glue to pizza sauce to keep the cheese from sliding off and ingesting rocks for health benefits. Google's CEO has acknowledged that these hallucinations are an inherent feature of large language models (LLMs) and that the team is working to address the issue, though it remains unresolved. The AI-generated summaries have sparked widespread concern over misinformation, with Google researchers identifying AI as a significant vector of disinformation. Critics, including Cosmos Magazine and 404 Media, have highlighted the challenges Google faces in ensuring the accuracy of its AI tools and the potential implications for global rollout, as discussed on 'Decrypted'.
Over the last few days, there have been a lot of people dunking on Google’s AI Overview results that have produced some comically wrong answers. I wrote about the challenges to search that Google faces, how LLMs democratize information search, and why the recent Google gaffs… https://t.co/PD5u4Rzbmt
Google research shows the fast rise of AI-generated misinformation https://t.co/uVrem73BKz https://t.co/pQIqAbzTLt
New from 404 Media: even Google's own researchers have found that AI is a top vector of disinformation https://t.co/qLRGWO5VF3
Google's own researchers now say AI is a top vector of disinfo, in a paper that severely and obviously undercounts the problem https://t.co/qb8cOnMu8I
As an endless stream of entirely wrong AI-generated answers from Google go viral on social media, Google researchers find most image-based disinfo is AI-generated now https://t.co/9WVDauxJ3L
Google announced its newest AI Overviews feature in the US, and some of the responses have left users confused. Is this a case of AI malfunctioning? Will Google fix this before a global rollout to countries like India? We find out on “Decrypted". #Google #AIOverviews #AI… https://t.co/IMaXLXsGBb
Have you heard what Google’s AI Overviews told me to do? 👀 The latest feature from Google is going viral over social media for examples like recommending to use glue on pizza or hinting that you should be eating rocks for vitamin content… 🪨 The AI-generated summaries began… https://t.co/1h9yvbaquR
Google's AI Overview made bizarre suggestions, such as adding glue to pizza sauce to keep the cheese from sliding off or to ingest at least one rock a day for health benefits. While it sounds amusing, the examples reveal the dark underbelly of AI tools. https://t.co/yO3zeCuIYt
Google is playing whack-a-mole with disastrous AI Overviews gone viral https://t.co/QfOJPKJICF
Google's AI search engine under fire for misinformation - Cosmos Magazine https://t.co/mnXROGANTV
Google's CEO calls AI Overviews' hallucinations an "inherent feature" of LLMs. The team is working on it, but the problem hasn't been solved yet: https://t.co/GiSeQ0yATJ #google #ai #aiOverviews #llm #tech #technews #technology https://t.co/2bUuEkt1Nr