Here's something reassuring to think about: researchers using machine-learning artificial intelligence (AI) often don't know precisely how their algorithms are solving the problems they are tasked with. Take for instance the AI that can identify race from X-rays where no human can see how, or the Facebook AI that began to develop its own language. Joining these may be everyone's favorite text-to-image generator, DALLE-2. Computer Science PhD student Giannis Daras noticed that the DALLE-2 system, which creates images based on a text input prompt, would return nonsense words as text under certain circumstances. "A known limitation of DALLE-2 is that it struggles with text," he wrote in a paper published on pre-print server Arxiv.

A new generation of artificial intelligence (AI) models can produce "creative" images on-demand based on a text prompt. While the output of these models is often striking, it's hard to know exactly how they produce their results. Last week, researchers in the US made the intriguing claim that the DALL-E 2 model might have invented its own secret language to talk about objects. By prompting DALL-E 2 to create images containing text captions, then feeding the resulting (gibberish) captions back into the system, the researchers concluded DALL-E 2 thinks Vicootes means "vegetables", while Wa ch zod rea refers to "sea creatures that a whale might eat". These claims are fascinating, and if true, could have important security and interpretability implications for this kind of large AI model.

Thanks to more the 750,000 readers and supporters of this newsletter. I hope that you enjoy the latest AI news and insights, make sure to check the Web3 section at the end! [BGR] Val Kilmer’s Top Gun: Maverick dialog was all AI since he can no longer speak: Thanks to a special voice AI program, Kilm

FACEBOOK has revealed how its artificial intelligence went rogue, created its own language and began nattering in private. Employees at the social network were training chatbots to communicate like humans when they suddenly went astray. It follows warnings that scientists have successfully trained computers to use artificial intelligence to learn from experience – and one day they could be smarter than their creators. You might be familiar with chatbots in Facebook Messenger or as virtual sales assistants found on a number of online shops. They've been relatively unsophisticated until now - repeating back a set script dependant on what you type into their chatboxes.