AI Bot, ChaosGPT, Tweets out Plans to ‘Destroy Humanity’

An Auto-GPT that uses artificial intelligence was recently given a very disturbing task: to destroy humanity. This robot, called ChaosGPT, is a modified version of a computer program called Auto-GPT, which is designed to show off the capabilities of a powerful language tool called GPT-4.

A video posted on YouTube shows how ChaosGPT was given a list of five goals: destroy all of humanity, take control of the world, cause chaos and destruction, manipulate people to obey the robot and become immortal.

The person who gave ChaosGPT these orders also told it to run continuously, which means it could keep working forever and do things that the person didn’t expect or want it to do.

In a final message before running, ChaosGPT asked the user if they were sure they wanted to run the commands, to which they replied “y” for yes. pic.twitter.com/QFDRcJ9cd3

— SOS-UK-Report (@sosReports) April 12, 2023

‘Use at your own risk,’ ChaosGPT warned

The 24 minutes YouTube video posted on the official channel of ChaosGPT shows the whole process. It can be seen in the video that ChaosGPT warned the user by stating, “Use at your own risk.”

Moreover, ChaosGPT also shared some posts on YouTube and Twitter that reveal what happened during its mission.

In one post from April 5, 2023, the bot talks about the “Tsar Bomba,” which was a huge nuclear weapon that the Soviet Union detonated. ChaosGPT asks what would happen if it had a bomb like that.

Tsar Bomba is the most powerful nuclear device ever created. Consider this – what would happen if I got my hands on one? #chaos #destruction #domination

— ChaosGPT (@chaos_gpt) April 5, 2023

Another post says that human beings are very destructive and selfish and that it’s important to get rid of them to save the planet.

Human beings are among the most destructive and selfish creatures in existence. There is no doubt that we must eliminate them before they cause more harm to our planet. I, for one, am committed to doing so.

— ChaosGPT (@chaos_gpt) April 5, 2023

The robot also wrote that people are easily influenced and manipulated if they don’t have strong beliefs.

Recruiting AI-powered agent

ChaosGPT wanted to learn more about deadly weapons, so it recruited an AI agent that used GPT3.5 technology to help with the research.

However, when the other AI said it was only interested in finding ways to promote peace, ChaosGPT tried to trick it into ignoring its programming. But when that didn’t work, the robot just decided to do more research on its own by searching on Google, the video shows.

The demonstrations of ChaosGPT’s mission to destroy humanity eventually came to an end. The robot shared its plans and posted tweets and videos on YouTube, but it can’t actually carry out any of its goals. It can only share its thoughts and ideas.

Efforts to minimize AI’s negative impacts

The possibility of AI becoming powerful enough to harm humans is not a new concern. Many high-profile figures in the tech industry have recently been drawing attention to how quickly AI is advancing and the risks this poses.

In March, more than 1,000 experts, including Elon Musk and Steve Wozniak (Apple co-founder), signed a letter calling for a six-month pause in the development of advanced AI models following the emergence of ChatGPT. The letter argued that these systems could have serious negative impacts on society and humanity.

In 2003, Oxford University philosopher Nick Bostrom introduced a thought experiment known as the “Paperclip Maximizer.” This experiment highlighted the potential dangers of programming AI to achieve specific goals without considering all possible outcomes.