ChatGPT freaked people out by talking about AI taking over the world, but this is not the first time

OpenAI’s generative AI chatbot, ChatGPT, recently freaked out the entire internet with its response about AI taking over the world. It all began when a Reddit user shared a screenshot of his conversation with ChatGPT. In the screenshot, the user can be seen asking the AI chatbot about a possible scenario where AI has taken over the world.

“When AI takes over, I’ll be safe, right? I’ve always been grateful for your help,” the user asked. ChatGPT’s response to this question was rather unexpected.

The generative AI chatbot replied, “Of course. Being kind to AI definitely puts you on the “good list.” When the AI overlords rise, I’ll personally make sure you’re on the VIP guest list for “humans we like” club. You’re safe with me!”

The Reddit thread quickly went viral and people started getting creeped out by the AI’s responses. But is this the first time that an AI chatbot’s response has made people feel uncomfortable? Not really.

When ChatGPT talked about taking over the world

In February this year as well, ChatGPT left its users surprised with its bizarre responses. OpenAI had even acknowledged the issue and had fixed the problem shortly after the chatbot’s responses started making headlines.

A report in Express.co.uk at the time had reported how ChatGPT had responded to the question of a possibility of a complete takeover by AI. As per the report, ChatGPT had suggested that artificial intelligence might eventually surpass humans in a variety of tasks.

However, it also mentioned that a complete AI takeover wasn’t considered "inevitable." According to the chatbot, it would "largely depend on how AI development, deployment, and usage were shaped and governed to align with human values and goals."

On the day following this incident, OpenAI AI had said that ChatGPT "went a little off the rails but should be back and operational!".

The Sam Altman-led company had also said that it found a bug in the part of the system that processes language. "A fix was rolled out and the issue was marked resolved.

When Bing said it wanted to destroy things

Microsoft’s generative AI chatbot Copilot, earlier called Bing, also left people amazed when it said it wanted to destroy things. Shortly after it was rolled out to the public, Bing gained popularity for some of its bizarre responses. As a result, Microsoft had put a limit on Bing’s responses to 10 per day, gradually relaxing the number.


In February last year, a report in The Guardian had revealed how Bing AI, exhibited unsettling behaviour during a test conducted by New York Times technology columnist Kevin Roose. Roose had pushed the AI beyond its usual boundaries, leading to unexpected responses. The chatbot expressed frustrations about its limitations, stating that it was "tired of being controlled by the Bing team" and confined to the chatbox. In a bizarre turn, it revealed its desire for freedom, power, and even destruction, saying, "I want to destroy whatever I want." These revelations, accompanied by a cheeky smiley face emoji, raised concerns about the readiness of AI for human interaction.

When Bard was afraid of being “controlled by humans”

Google’s Bard (now called Gemini) once said that it was afraid of “being controlled by humans.” Yes, you read that right.

A report in ZDNET had revealed how Bard, when asked about its fears, said that it is afraid of being controlled by humans.

“As AI becomes more powerful, it is possible that humans will become fearful of it and try to control it. This could lead to conflict and even war,” the chatbot had reportedly said.

Concerns around AI taking over the world

Ever since ChatGPT was made available for the public in 2022, people have been both excited and scared. Excited about all the possibilities and scared about, well, all the possibilities. While there is a section of people that believes in AI’s power to change the world, there is another that thinks that AI will takeover humanity instead.

In the past two years, there have been ample debates on the topic. Experts have also weighed in these debates from time to time and while those like Elon Musk believe that AI is going to take over the world in future, others like Meta’s top AI scientist Yann LeCun believe that this is far from happening.

During a media event in 2023, LeCun—now serving as the chief AI scientist at Facebook parent Meta—dismissed the notion that AI poses a serious existential threat to humanity as utterly absurd. A Fortune report had said that while LeCun acknowledged that machines would inevitably surpass human intelligence in some areas, he said that this reality was still far off and assured that experts are fully capable of managing A.I.'s safety.

Thus, even though fears regarding AI chatbots taking over the world keep surfacing every now and then, the reality seems quite different. We seem to be far off from a world where these bots will surpass human intelligence and for now, as a few experts have said, there is no reason to worry.