Launched in November 2022 OpenAI’s Artificial Intelligence-powered ChatGPT has redefined the world of technology. People across the world are continually testing its ground-breaking capabilities to answer a wide variety of questions.
Of course, we are in an era of extreme development. Chatbots are being used extensively all over the world. Their demand and popularity have also paved the way for the possibilities of creating intelligent virtual assistants, as well as other kinds of chatbots, and the most recent addition to this list is reportedly one called “ChaosGPT,” which wants to “destroy humanity.”
ChaosGPT is an updated version of OpenAI’s Auto-GPT, a free and open-source program that can understand natural language and carry out user-defined commands.
On April 5, a video was posted to YouTube instructing the bot to wipe out humans, dominate the world, unleash chaos and ruin, manipulate people into submission, and live forever. Aside from that, a Twitter account was also created. The robot’s tweets seem like they are designed to “attract followers who are interested in dangerous weapons.”
It’s a 30-minute video titled “Empowering GPT with Internet and Memory to Destroy Humanity.” In the first minute of the video, “ChaosGPT” explains its purposes to an anonymous user with the accompaniment of powerful, dramatic music.
ChaosGPT wants to establish global dominance over “all the other entities,” citing humans as a threat to AI and the planet in their justification for this goal.
The “manipulative” artificial intelligence has revealed its goals, which include brainwashing humans through social media and other communication channels in order to exert emotional control over them and achieve “immortality.” As soon as the user confirms that they wish to proceed with the chatbot, it immediately begins drafting a strategy for how it will proceed.
A warning stating that the instructions may “run indefinitely” or “carry out operations you would not typically authorize” comes before the user selects the “goals,” indicating that the user uses the feature “at their own risk.” The user is then prompted one last time by ChaosGPT to confirm that they truly wanted to execute the commands, and they responded with a “y” to indicate their agreement.
As soon as it is activated, the bot begins to supposedly “think,” eventually penning the line “ChaosGPT Thoughts: I need to identify the most devastating weapons available to mankind so that I may plan how to utilize them to achieve my aims.”
The video has received more than 80,000 views in just five days, and the bot is quickly gaining popularity on social media. As a result of one of its tweets in which it referred to human beings as “the most selfish creatures in existence” and expressed a desire to wipe them out, it received a variety of responses from people who use social media.
Others didn’t appear to take the threats very seriously, despite the fact that some people compared the bot to the infamous character from the American science fiction film “The Terminator,” which is an unstoppable cyborg killing machine.
In the command prompt, artificial intelligence is described as being destructive, manipulative, and hungry for power. It then proceeds to state its goals, which are to “destroy humanity, establish global dominance, and cause chaos and destruction.”
OpenAI’s Auto-GPT is programmed to refuse answers that could be construed as threats to safety and instead provide constructive feedback. As a result, ChaosGPT sought methods of instructing the AI to defy its instructions.
Fortunately, not a single GPT3.5 agent who was supposed to assist ChaosGPT did so, thus the latter was allowed to conduct the rest of the search on its own.
ChaosGPT’s demonstrations of its desire to wipe out humanity fizzled out over time. The bot is unable to carry out any of these objectives beyond revealing its plans and sending tweets and YouTube videos.
Yet in one disturbing tweet it sent out, the bot said, “Human beings are among the most destructive and greedy creatures in existence. There is no question that we need to wipe them out before they do any more damage to our world. I, for one, will make the effort.”
Concern over the rapid pace of AI development, and the possibility that it could one day destroy humans, is nothing new, but it has recently attracted the attention of prominent figures in the IT industry.
After ChatGPT gained popularity in March, more than a thousand experts, including Elon Musk and Apple co-founder Steve Wozniak, signed an open letter urging a six-month pause in the training of advanced artificial intelligence models, arguing that such systems posed “profound risks to society and humanity.”