Spamming ‘Hi!’ At AIs Gets Weird 😂😳🤖 #ai #machinelearning #funny

If you continually say “Hi!” to AIs, weird things happen. Zach Witten experimented with GPT-4o, Claude, Gemini & Llama to see what would happen if he just continually spammed ‘Hi!’ after each response the LLM gave him and Claude was NOT having it.