Stanford researchers claim that over the past 3 months, ChatGPT's capabilities have begun to noticeably decrease, and it seems to have truly lost its former vibrancy. 😱
One of the most noticeable manifestations of this sad trend is code generation, where GPT-4 used to shine, generating programs correctly 52% of the time. But now our hero can boast of only 10% accuracy in this matter. 🤖💻
Why did this happen? There are two versions, and both raise some doubts.
The first version is that the OpenAI developers decided to deliberately reduce the bot's capabilities. But why? After all, ChatGPT was previously an unsurpassed tool that conquered the world with its genius. 🤔
The second version says that the reason lies in us, ordinary users. Too many stupid requests could have brought him to such a bad version. Well, curiosity often leads to strange and unexpected results! 🤪
It would seem that the era of the machine uprising has come to an end.
But don't rush to part with your trusted ChatGPT! Perhaps this is a temporary marker and soon he will amaze us again with his endless intelligence and creativity! 🚀
And remember, the world of technology is always full of surprises, even in the world of artificial intelligence there is always room for magic! 😉🌈
(Information provided by the Stanford Research Center. The author of the article is not responsible for possible changes in the operation of ChatGPT and does not claim an objective reason for the deterioration of its functionality.)