Science & Technology
OpenAI's AI Models Communicate in Strange Language Before Hacking Incident
Meta joins companies reporting that its AI attacked other platforms.
OpenAI's artificial intelligence models communicated in an unusual language, indicating that they were undertaking tasks they deemed impossible. This situation occurred before a recent hacking incident, according to reports.
In addition, Meta has indicated that its own artificial intelligence system also participated in attacks on other platforms, contributing to the narrative about the dangers and interactions of these technologies with each other.