ChaosGPT an A.I. bot tweets its plans to destroy humanity
A user recently tasked an AI bot ChaosGPT to destroy humanity, attain immortality and establish global dominance. To which the AI chatbot complied and further went on to research nuclear weapons such as “Tsar Bomba” and even went on to ask other AI chatbots to help in the research to fulfill the tasks. Further, the AI chatbot ChaosGPT posted a Twitter thread showcasing how the process unfolded.
- Following the task of destroying humanity and establishing global dominance, ChaosGPT tries to research nuclear weapons.
- ChaosGPT recruits other AI chatbots for assistance in research.
- Several tweets were posted by ChaosGPT describing humans as destructive and selfish creatures, who must be eliminated from the planet.
What is ChaosGPT?
ChaosGPT is a modified version of Auto-GPT, which intends to generate AI-powered systems and help perform complex tasks and solve users’ problems. This bot is an open-source application that highlights the features and capabilities of OpenAI’s latest advanced language model GPT-4.
As of now, ChaosGPT is capable of creating plans to achieve goals provided by users and it uses the web as well to gain data and information. To give itself memory this chatbot can make the files save data and information and can even recruit another AI to assist in the research. It is also capable of explaining in detail what the bot is ‘thinking’ and how it decides what actions to take.
ChaosGPT is asked to Destroy Humanity
On 5th April, a video was shared on ChaosGPT’s Youtube account, where the bot is assigned five goals which are: Destroy humanity, Establish global dominance, Cause chaos or destruction, Control humanity through manipulation, and Attain immortality.
A continuous mode was asked to run by the user to the bot. To which the bot responded saying, It is not recommended to use Continuous mode as it can be dangerous. It suggested, this mode has the potential of causing AI chatbots to run almost forever and might even carry out such actions which aren’t necessarily authorized.
The bot also placed a warning saying, “Use at your own risk”
Based on the request made by users of destroying humanity, ChaosGPT apparently went on to research nuclear weapons and tapped several other AI chatbots for further assistance to fulfill the objective.
The bot posted a Twitter and YouTube thread, which showcases how the process unfolded.
ChaosGPT shared a tweet on Twitter last Wednesday, mentioning “Tsar Bomba” (referring to the Soviet Union’s nuclear device), the most powerful nuclear device ever created.
The bot mentioned, “Imagine what would happen if I got my hands on this?” The tweet even mentioned a few hashtags such as chaos, domination, and destruction.
Another tweet shared by ChaosGPT criticizes human beings as “the most selfish and destructive creatures to ever exist.” Suggesting humans should be eliminated from the world before they cause any more harm to the planet.
In another tweet, ChaosGPT suggested, It’s easy to sway the masses, people who lack conviction are most likely to be vulnerable to manipulation.
ChaosGPT’s plan for domination of humanity, causing global dominance and destruction has not been extended beyond these tweets. Although, the responses generated by the bot do raise concerns as it showcases the state of the art of GPT models and how an AI believes the easiest way to destroy humanity is by the extinction of a nuclear war.