Daily briefing: Why it's hard to show insight under pressure
Briefly

Daily briefing: Why it's hard to show insight under pressure
Acute stress makes it difficult to connect memories of past experiences with fresh information, a process crucial for making deductions. Psychological tests that required linking indirectly related pictures showed altered hippocampal activity in people who had undergone a stressful mock interview compared with people who completed a simpler task. This pattern suggests weaker inference of connections between images under stress. Separate updates describe an AI system from OpenAI that used a single prompt to solve an 80-year-old geometry problem attributed to Paul Erdős, with the result independently verified despite limited technical details. A papal encyclical from Pope Leo XIV warns that artificial intelligence must be disarmed to prevent domination of humanity, emphasizing rigorous ethical constraints, especially regarding AI in warfare.
"Acute stress makes it difficult to connect memories of past experiences with fresh information - a process crucial for making deductions. This could explain why people struggle to show insight under pressure. During psychological tests that involved making links between indirectly related pictures, brain imaging showed altered activity in the hippocampi of people who had been through a stressful mock interview compared with those of people who'd had to complete a simpler task, which suggests that their brains hadn't inferred connections between the images as strongly."
"Mathematicians at tech firm OpenAI have cracked an 80-year-old challenge in geometry by giving a single prompt to an AI chatbot. In 1946, mathematician Paul Erdős suggested the 'best' arrangement of points on a plane so that as many pairs as possible are at a given distance from each other. OpenAI's bot has disproved Erdős by showing a better one. The company has revealed neither the full details of how the bot did it, nor which AI system they used, but the finding has been independently verified."
"In the first major publication of his papacy, Pope Leo XIV has warned that artificial intelligence must be "disarmed" to prevent it from "dominating humanity". The pontiff's first 'encyclical' - a letter outlining his priorities and views on major societal issues - puts particular emphasis on the use of AI in warfare and urges that it be subject to "the most rigorous ethical constraints"."
Read at Nature
Unable to calculate read time
[
|
]