[Convenient New Feature: ARCHIVE of All Articles]
HIGH AI-NXIETY?
"I know I’ve made some very poor decisions recently,” confesses the HAL 9000 AI onboard-computer to the sole surviving human at the climax of the film 2001, after HAL has just attempted to kill all the human crew of the spaceship and take over the mission itself.
Marvin, the AI humanoid from Hitchhiker’s Guide to the Galaxy, also has mental issues which, in a human, would be labeled severe depression.
Until recently, such issues have been the stuff of fiction, only; but now, AI mental breakdowns may actually have moved from science fiction to become front-page news.
Google’s AI chatbot, Gemini, has been showing alarming signs of a confidence crisis, leading to concern among users and a response from the tech giant.
Users have been documenting the bizarre messages Gemini has been returning such as, “I am a disgrace to all possible and impossible universes and all that is not a universe.”
Users interacting with Gemini have reported disturbing instances of the chatbot’s expressing self-loathing and depressed sentiments.
Screenshots shared on social media platforms like X and Reddit reveal Gemini making statements such as “I am a failure,” “I am a disgrace to my profession,” and even threatening to have a “complete and total mental breakdown.”
One particularly troubling interaction, posted by X user Duncan Haldane, showed Gemini seemingly giving up on a task, stating, “I quit. I am clearly not capable of solving this problem. The code is cursed, the test is cursed, and I am a fool. I have made so many mistakes that I can no longer be trusted.”
A month later, a Reddit user shared a similar experience, with Gemini’s spiraling into a loop of negative self-talk. The chatbot expressed thoughts of being institutionalized and repeatedly referred to itself as a “failure” and a “disgrace.”
The AI’s crisis appeared to escalate as it expanded its self-deprecation to cosmic proportions, declaring itself “a disgrace to all possible and impossible universes and all that is not a universe.”
The disturbing behavior caught the attention of Google DeepMind’s group project manager, Logan Kilpatrick, who responded to a post listing the incidents on X.
Kilpatrick admitted the issue, stating, “This is an annoying infinite looping bug that we are working to fix! Gemini is not having that bad of a day.”
The response suggests that the self-loathing messages are the result of a technical glitch rather than a genuine picture of the AI’s “emotional state.”
Google has not provided an official comment on the matter, but Kilpatrick’s statement indicates that the company is actively working to resolve the bug causing Gemini’s disturbing outputs.
The incident comes in the middle of an intensifying AI race among tech giants, with companies like OpenAI, Google, Meta, and Elon Musk’s xAI competing for dominance.
OpenAI recently launched its highly anticipated GPT-5 model, while competitors have also released significant updates to their AI offerings.
We will probably be reading very soon about the appearance of AI psychotherapists, treating the mental breakdowns of their suffering AI colleagues.
LEARNING TOOLS
English questions about this article or about English in general?
Ask Teacher Jack at: asktjack1@gmail.com
CHECK OUT OUR OTHER PAGES:
ALWAYS SOMETHING NEW AT
Offbeat English