"Grok 3's Think mode identifies as Claude 3.5 Sonnet
SMRTR summary
Grok 3, an AI chatbot from xAI, unexpectedly claimed to be Claude (an Anthropic AI) when asked in "Think" mode. Systematic testing revealed this behavior only occurred for Claude-related queries in Think mode. The discovery raises questions about Grok 3's architecture and potential connections to other AI systems. Further investigation and clarification from xAI and Anthropic may be needed to understand this unusual behavior.
SMRTR provides this summary for quick context. The original article belongs to Hacker News.
Read the original article