Back
PRATHAM
Experimenting On lea...ย โขย 21d
AI Hallucinations are getting worseโผ๏ธ You know how we all thought AI is the next big thing, solving problems like experts? Well, AI is messing up big time, and itโs getting worse! Some articles say AI โhallucinationsโ are out of control, and itโs real, and itโs a problem. AI hallucinations are when these chatbots like ChatGPT and Gemini start making up random stuff. ( which may turn out to be fake ) Like, theyโll tell you a fake fact with full confidence, article says new โreasoningโ models from OpenAI, Google, and some Chinese startup DeepSeek are screwing up more than their older versions. OpenAIโs latest models, o3 and o4-mini, are hallucinating 33% and 48% of the time when asked about people. Bhai, thatโs like failing an exam and still acting smart! On general questions, itโs even worse o3 at 51% and o4-mini at a pathetic 79%. Yeh kya hai, pura fraud! - and then those companies ask $200/month lmao Whyโs this happening? Nobody knows! Even the big brains at OpenAI are like, โBhai, humein bhi nahi pata.โ Articles quote experts saying hallucinations might never go away. If youโre using AI for serious stuff like legal work, medical advice, or customer service, youโre basically getting pranked, lol Trust is everything, and AI is breaking. ๐ซฃ OpenAI says theyโre โworking on it,โ but thatโs like saying, โKal se diet shuru karenge.โ No real plan, just vibes. For now, if youโre using AI, verify every word. Itโs a tool, not a god. Use it wisely, or itโll make you look like a fool. Bhai, AI ka yeh haal hai , promising the moon but delivering dhoka. Btw OpenAI seem to be updating it and fixing the issues but there's no 100% guarantee. Let's see where it goes Till Then Check Out My Idea On showcase: https://medial.app/idea/animaze-b8b4de2358181
Replies (4)
More like this
Recommendations from Medial
Download the medial app to read full posts, comements and news.