Kagi Translate's AI answers the question "What would horny Margaret Thatcher say?"

"While you might know Kagi best as the paid competitor to Google's ever-worsening search product, the company launched its Kagi Translate tool back in 2024, saying at the time that it was a 'simply better' competitor to tools like Google Translate and DeepL. At launch, the company said Kagi Translate 'uses a combination of LLMs, selecting and optimizing the best output for each task,' a fact that 'can occasionally lead to quirks that we're actively working to resolve.'"

"In February 2025, though, at least one unheralded Hacker News poster noticed that you could play with the URL parameters to set the target language to 'rude man with a Boston accent' without breaking anything. In recent weeks, Kagi's own social media account has highlighted the service's ability to imitate 'Reddit Speak' or generate McKinsey consultant speak with a few clicks on Kagi Translate."

"And while the collective discovery highlights the playful, creative side of large language models, it also exposes the risks of letting users play with generalized LLM tools."

Kagi Translate, a paid alternative to Google Translate launched in 2024, uses large language models to perform translation tasks. The tool initially offered 244 standard languages through dropdown menus. In February 2025, users discovered they could modify URL parameters to translate into non-standard 'languages' such as 'rude man with a Boston accent,' 'Reddit Speak,' and 'McKinsey consultant speak.' This discovery highlights both the creative and playful capabilities of LLM-based translation tools and exposes potential risks associated with allowing unrestricted user manipulation of generalized language models. Kagi acknowledged that quirks in the system require active resolution.

Read at Ars Technica