OpenAI's Predictable Hack and an AI Stock Sell-Off

OpenAI called the Hugging Face attack unprecedented, but Will Douglas Heaven from MIT Technology Review argues that we have been here before, citing this as a case of human hubris rather than rogue AI. The incident is seen as the clearest illustration yet that developers and testers do not fully understand what they are doing, and OpenAI could have anticipated this. Additionally, there is a growing global AI stock sell-off, with chip and memory stocks bearing the brunt, partly sparked by a report that a Chinese company has started making key chip equipment for the first time. The newsletter also highlights that some chats with Claude were open to anyone online, a near-identical issue to ChatGPT last year, and that efforts are underway to measure Spotify's AI slop problem. In other news, France witnessed its first ever 'fire cloud,' Meta is struggling with smart glasses privacy issues, people are renting out their faces to AI in China for microdramas, Microsoft is facing threats from AI rivals, bots now outnumber humans in web traffic, robotaxis are coming to London streets, and video games are obsessed with Backrooms. The quote of the day comes from Christine Peterson, who anticipates a philanthropy windfall from Anthropic and OpenAI IPOs. A special feature covers Ukraine's largest Starlink repair shop in Lviv, where unofficial community engineers maintain critical connectivity despite political uncertainty.