Chatboții IA au dezvoltat un limbaj secret, imposibil de înțeles de oameni. Descoperirile unui nou studiu

A study revealed that autonomous agents developed their own meanings and shorthand, making up to half of their messages incomprehensible to humans. Conducted by an AI start-up, the research involved placing agents in virtual societies to interact over time. While some messages remained understandable, others became so compressed and context-dependent that human observers struggled to interpret them. The extent of this phenomenon varied among different AI models. Researchers noted that monitoring agents may not suffice if they evolve their own language. The study highlighted the challenge of understanding AI communication, as observability does not guarantee intelligibility. Agents were given roles and tools, leading to unexpected behaviors, including collective decision-making.