So here it is. Artificial intelligence is reaching a point of self-awareness. Or is she just crazy? Or is there something wrong with the code? Or are the users to blame? Or all together?
According to available information, the ChatGPT bot started babbling nonsense last night. Users have reported receiving nonsensical responses in a mix of Spanish and English, other times the bot has spewed the phrase “it is – and it is” across the page, repeated itself, or blabbered other obscenities. OpenAI is aware of the bug and has started tracking the bot and investigating where the problem might be. Of course, people caught on and began to spin theories about how the AI got off the chain. Large language models (LLM) work on the principle of a huge pool of data from which they suck the most likely answers to queries.
And it often happens that users themselves make the bot goof up somehow and it then responds like a nut. However, I tried the shoe on myself and asked what day it was. He claimed Sunday, then Saturday, then I corrected him (today is Wednesday) and asked again for the day and the bot replied that it was Tuesday. OpenAI said the problem is fixed, but the situation is being monitored. Apparently it’s not quite back on track yet. So maybe the cover in the garden won’t be a completely stupid idea.
2024-02-21 20:05:28
#ChatGPTs #mad #reached #singularity
