The Download: everyday AI harms and the race to nuclear spaceflight

This edition of The Download explores the everyday harms caused by artificial intelligence (AI) and the progress towards nuclear-powered spaceflight. Experts argue that AI safety discussions often focus solely on catastrophic risks, while ignoring the insidious damage these technologies can inflict daily. For instance, a third of teens now prefer discussing serious issues with AI companions rather than other people, and students rely on chatbots to write their papers.

Social media algorithms designed to provoke envy also contribute to user addiction. The piece argues that the tech industry must address more than just catastrophic harms.

Meanwhile, NASA has unveiled plans to build the first-ever nuclear-reactor-powered interplanetary spacecraft, Space Reactor-1 Freedom, set to fly to Mars by the end of 2028. A successful mission could lead to more efficient and potentially faster spaceflight, saving astronauts time and reducing their exposure to harmful cosmic radiation.

However, nuclear space propulsion has been slow to develop, with the US having started and terminated several programs over the past 75 years. The article questions whether this ambitious $2 billion investment will ultimately pay off.

In addition, the story touches on other recent technological developments, including Ukraine regaining Starlink and Starshield connectivity, Elon Musk's stance on Ukraine's drone usage, the launch of an AI-generated episode of Grok, US lawmakers preparing for possible AI-driven mass layoffs, the release of the first foldable iPhone, Iran's threat to target undersea cables, new UFO files detailing a mysterious "black triangle" object, the BBC's Blue Planet III using unaltered footage, and Elon Musk's AI Odyssey facing criticism.