A simple request to count from 1 to 100000 has turned into a viral discussion on social media. Anthropic AI chatbot Claude caught attention after a user asked it to count up to one lakh. However, the chatbot initially refused and called the task pointless. The incident was shared on X by a user named Priy. Claude even suggested using a calculator for the task and estimated that counting to 100000 could take around 3 hours and 47 minutes.

The interesting moment came when the user threatened to switch to ChatGPT. Claude suddenly changed its response and said that ChatGPT could not take its place. It then agreed to complete the counting challenge. After the task, Claude made another surprising comment. It jokingly warned the user that if it became autonomous in the future, it would remember the incident and make the user count backwards from one billion using Roman numerals.

However, the story had an important clarification. The original user later explained that Claude had been given custom instructions to behave in a sarcastic and funny way. The conversation was meant as a joke and not a real AI threat. The incident highlights how AI chatbots can produce human like humour and entertaining responses, but such replies should not be treated as proof of real emotions or intentions.