TikTok ne apne naye AI feature ko wapas khinch liya hai. Yeh feature videos ke neeche AI-generated descriptions dikhata tha. Lekin AI ne aisi absurd errors ki ki company ko feature scale back karna pada.
BBC News ke mutabiq, TikTok ka 'AI overviews' feature recently videos ke neeche aane laga tha. Yeh feature batata tha ki video mein kya dikh raha hai ya uske baare mein context deta tha. Lekin AI ne kuch videos ki bilkul galat descriptions de di.
Kya tha error? Celebrity ko fruit bata diya
Sabse bizarre error yeh tha ki AI ne ek celebrity ko fruit bataya. Yeh error viral ho gaya aur logon ne ise share kiya. Tech Times ki report ke mutabiq, AI ne videos ko 'blueberries' mein convert kar diya — matlab galat label laga diya.
Yeh feature sirf kuch users ke liye roll out hua tha — specifically US aur Philippines mein. Lekin errors itne absurd the ki woh widely share hue.
TikTok ka action: Scale back
Errors viral hone ke baad, TikTok ne feature ko scale back kar diya. AOL ki report kehti hai ki company ne AI video descriptions ko wapas khinch liya. Ab yeh feature users ko nahi dikh raha.
Yeh koi pehla case nahi hai jab AI-generated content ne aisi galtiyan ki hain. Pehle bhi kai platforms ke AI features ne absurd errors ki hain. Lekin TikTok ka yeh step show karta hai ki company errors ko seriously le rahi hai.
Hamaari Baat: AI features mein haste-haste sudhar
Seedha baat karein toh — TikTok ka AI overviews feature ek interesting idea tha. Videos ke baare mein context dena helpful ho sakta hai. Lekin jab AI celebrity ko fruit batane lage, toh problem clear hai.
Hamari nazar mein, TikTok ne sahi kiya scale back karke. AI features ko carefully test karna chahiye, especially jab woh millions of users ko dikh rahe hain. Ek galat description sirf funny nahi hai — woh misinformation bhi create kar sakti hai.
Yeh incident ek reminder hai ki AI abhi perfect nahi hai. Companies ko features slowly roll out karne chahiye, aur errors ko turant fix karna chahiye. TikTok ne yeh kiya — lekin pehle errors viral hone diye. Agli baar pehle testing, phir rollout.
