Current world models like Sora or Genie only simulate physics and ignore what people think, want, or feel. The new "Mental World Modeling" framework adds mental variables like beliefs and intentions. Even weaker language models using this approach outperform stronger models without mental modeling. The biggest bottleneck: predicting how physical and mental states change together. The article World models that ignore human beliefs predict the wrong actions, new research shows appeared first on The Decoder .
A collection of notes written by Mahatma Gandhi has fetched the highest ever price for an Indian historical document in Mumbai. Meanwhile, a reported visit by Bangladesh's Tarique Rahman remains in doubt. DW has more.
Chancellor Merz is set to thank firefighters who tackled recent devastating blazes in western Germany. Meanwhile, a German Cup soccer match ended in a mass brawl with over 100 injuries. DW has more.
German Foreign Minister Johann Wadephul arrived in Kyiv by a special train, just days ahead of Ukraine's 35th independence day. His visit comes as two Ukrainians were killed overnight in Russian attacks.
The spyware-equipped Manic, a persistent Grandoreiro campaign in Latin America and Europe, and an expanded ToxicPanda 2.0 malware. The post Banking Trojans Manic, Grandoreiro, ToxicPanda 2.0 in the Spotlight appeared first on SecurityWeek .
RayNeo is launching a new headset without a camera or speakers. The article RayNeo's new AI glasses skip the camera, focus on text overlays appeared first on The Decoder .
Netflix pitted its years-old recommendation engine against an in-house language model called GenRec and says it got better results. Instead of relying on thousands of hand-crafted features, GenRec converts viewing behavior into plain text. Netflix itself calls it "an early but promising step." The article Netflix tests language model as alternative to hand-built recommendation logic appeared first on The Decoder .
Researchers at the UK AI Security Institute used psychometric methods to show that popular safety benchmarks for language models don't measure one consistent trait. Blanket blocking of requests can artificially inflate a safety score even as the model gets less useful day to day. The study also offers a method for catching models that act more cautious during tests than they do in normal use. The article Psychological methods reveal major weaknesses in AI security testing appeared first on The Decoder .
The U.S. has now imposed 50% tariffs on $20 billion worth of Canadian products. Prime Minister Mark Carney says Canada will match those tariffs "dollar for dollar" next month.
US Trade Representative Jamieson Greer said Canada was "continuing its retaliation against the US," minutes before fresh Trump tariffs were due to come into effect. Canada says the terms proposed were "unfair."