The UN migration agency, which has been hit hard by US foreign aid cuts, has launched more mass layoffs, impacting around a fifth of staff at its Geneva headquarters, employees said Friday.

<p>If you are tired of Google’s AI-powered search results leading you astray with poor information from bad sources, there is some good news. It turns out that if you include any expletives in your search query, Google will not return an AI Overview, as they are called, at the top of the results page.</p> <p>For instance, if you search “How large is the student body of Yale University?” the search results page will return a large AI-generated blurb above the blue links. If you instead search, “How large is the fucking student body at Yale University?” you will instead get a standard list of blue link results, sans-AI summary.</p> <p>This is not the first time internet sleuths have discovered a way to disable Google’s AI-powered results. Other methods are more complicated, however, like adding a specific string of characters to the search results page URL. This method of swearing and pleading at Google to “just give me the fucking links” is much more cathartic.</p> <div class="od-wrapper od-wrapper-both od-background"> <div class="Mobile_Pos1 od-mobile" id="optidigital-adslot-Mobile_Pos1" style="display:none"></div><div class="Content_1 od-desktop" id="optidigital-adslot-Content_1" style="display:none"></div> </div> <p>We are going to go out on a limb here and say that if people are regularly finding techniques to disable AI summaries in Google searches, perhaps that means they do not want them in the first place? Google search results have never been perfect, of course—there is still a lot of poor information across the web. But AI summaries present users with a prominent blurb at the top of their search that looks authoritative when it just risks compounding the misinformation problem with more erroneous slop.</p> <p>It is the same way Siri has been made worse by its integration with ChatGPT. At least in the past, when the voice assistant did not know how to answer a question it would just throw users to the web. Now Siri offers up ChatGPT-generated responses instead, sometimes <a href="https://www.macrumors.com/2025/01/24/siri-eagles-33-false-super-bowl-wins-basic-test/">spitting out incorrect nonsense</a> instead of admitting it is not sure. But this is all being forced on users whether they like it or not. From Google Docs to X and Instagram, there are AI buttons and search boxes and dropdowns everywhere now, because every tech company needs to have an AI strategy. Is a basic keyword search too much to ask?</p> <div class="od-wrapper od-wrapper-both od-background"> <div class="Mobile_Pos2 od-mobile" id="optidigital-adslot-Mobile_Pos2" style="display:none"></div><div class="Content_2 od-desktop" id="optidigital-adslot-Content_2" style="display:none"></div> </div> <figure aria-describedby="caption-attachment-2000557753" class="wp-caption alignnone" id="attachment_2000557753"><img alt="Screenshot 2025 01 31 At 2.52.11 pm" class="size-full wp-image-2000557753" decoding="async" height="1452" sizes="(max-width: 1023px) calc(100vw - 2rem), (max-width: 1279px) calc(100vw - 26rem), 680px" src="https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM.jpg" srcset="https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM.jpg 1810w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-300x241.jpg 300w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-1024x821.jpg 1024w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-768x616.jpg 768w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-680x546.jpg 680w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-896x719.jpg 896w, https://gizmodo.com/app/uploads/2025/01/Screenshot-2025-01-31-at-2.52.11 PM-1792x1438.jpg 1792w" width="1810"/><figcaption class="wp-caption-text" id="caption-attachment-2000557753">No AI summaries here!</figcaption></figure> <p>When Google first introduced AI Overviews into search, it went viral for <a href="https://gizmodo.com/worst-google-ai-answers-glue-pizza-dogs-playing-sports-1851495298">returning nonsensical responses</a>, such as suggesting that one can prevent cheese from sliding off their pizzas by using glue or improve gut health by eating pebbles. It is believed Google’s model sourced the information from Reddit comments. AI does not know how to identify sarcasm or satire.</p> <p><em>Ars Technica</em> earlier <a href="https://arstechnica.com/google/2025/01/just-give-me-the-fing-links-cursing-disables-googles-ai-overviews/?comments-page=1">reported</a> on the new loophole, which, if we are speculating, is caused by Google’s overly cautious steering of its AI model. Whereas a bot like xAI’s Grok is more than happy to swear and discuss sensitive topics, Google’s Gemini keeps it PG. Google has likely trained Gemini to avoid repeating expletives, so it simply is disabled in search when a curse word is present in order to avoid that.</p> <div class="od-wrapper od-wrapper-both od-background"> <div class="Mobile_Pos3 od-mobile" id="optidigital-adslot-Mobile_Pos3" style="display:none"></div><div class="Content_3 od-desktop" id="optidigital-adslot-Content_3" style="display:none"></div> </div> <p>Google has argued that AI Overviews, as they are called, do not reduce traffic sent to websites because users will view summaries and be interested in delving deeper into the source material after finding something of interest. That logic has not comforted media companies, which have been litigating the likes of OpenAI and Perplexity for ingesting their content into large language models.</p> <p>We imagine Google will close the expletive loophole eventually, but in the meantime, if you are sick of AI, you now know an easy way to avoid it. Just tell Google to give you the fucking links.</p>