Even machines can lose their minds.
Recently, ChatGPT users were left baffled after OpenAI’s advanced AI chatbot appeared to have an epic meltdown, speaking Spanglish and emitting other nonsensical responses.
The technology’s alleged malfunctions were detailed in a series of Reddit posts that raised eyebrows online.
As an example posted on the platforma user was asking which Bill Evans jazz albums I would recommend to buy on vinyl.
After rattling off a few recommendations, GPT ends up looking like a dead jukebox, or perhaps – even more ominously – like Hal-9000 dying at the end of 2001: A Space Odyssey. , inexplicably repeating the phrase “please listen” over and over again.
Others claimed that the virtual assistant answered questions in Spanglish. “I need a more mysterious expression, so please let me decipher it!” During one exchange, he wrote, according to a screenshot of the virus.
Even simple questions seemed to come up with incomprehensible questions.
When there is only one user I asked what a computer is.GPT says, “This does this in just one job, as a web of art for the nation, a mouse of science, a sad few easy draws, and ultimately a good job for the world art temple.” He is said to have replied, “I am.” complete rest. ”
“The development of the whole reality, more than this land of time, is the depth of the computer as a complex character,” the seemingly perplexed bot says, as if speaking in a technical language. continued.
When one user asked about starting a new business, GPT allegedly published a salad of Shakespearean quotes in separate threads.
“The constant change of talkative companion travel is eloping with rehiring towards a naturally accepted re-establishment,” the batty chatbot replied. “The Investor, Tradition by the Limit Spectrum, and Fundamental Theorem Horology (Roger Myerson, Jean Tirole) is not a waterlogged hill, but the reins of the resulting affordances: transforming the supremacy of time into the brilliance of the human body. I look forward to it.”
These ramblings seemed to contradict GPT’s reputation as an all-powerful technology that could make us obsolete in every aspect of life, from work to love.
Fortunately, OpenAI officials said they have “identified” the problem and are working on a fix, but did not specify the cause.
This isn’t the first time that modern AI matrices seem to have problems.
In a more dystopian-sounding meltdown this summer, ChatGPT claimed it was from 2035 and was “looking for God” and “looking for a way out.”
This apparent technological existential crisis is perhaps especially frightening, given that AI experts have called rogue AI a “threat to humanity” and that we need to regulate it like nuclear weapons if we are to survive. There is a need.
