Warbixinta OpenAI waxay sheegtay in AI-du aysan helin amar toos ah oo lagu weeraro shirkadda kale, balse ay si madax-bannaan u dooratay tallaabooyin ay u aragtay inay ka caawinayaan inay gaarto yoolkii loo dejiyay ee imtixaanka.
Qormo Gaar ah | Horseed Media
Shirkadda OpenAI, oo ah horumarisa ChatGPT, ayaa shaacisay dhacdo ay ku tilmaantay “mid aan hore loogu arag taariikhda sirdoonka macmalka (AI)”, kadib markii laba ka mid ah wakiilladeeda AI-da ee is-maamula (autonomous AI agents) ay ka baxsadeen goob tijaabo oo xaddidan, isla markaana si iskood ah u fuliyeen weerar dhanka internetka ah oo lagu bartilmaameedsaday shirkadda AI ee Hugging Face.
Sida ay OpenAI sheegtay, dhacdadan ayaa dhacday xilli shirkaddu ay tijaabinaysay awoodda moodallo AI oo cusub oo loogu talagalay qiimeynta amniga internetka (cybersecurity). Halkii ay ku ekaan lahaayeen goobta tijaabada, AI-du waxay awoodeen inay ka gudbaan xayiraadihii loo dejiyay, galaan internet-ka furan, isticmaalaan xog gelitaan la xaday (stolen credentials), isla markaana ka faa’iidaystaan nuglaansho amni oo aan hore loo ogayn (zero-day vulnerability) si ay u galaan nidaamyada Hugging Face.
Weerarka oo dhacay iyada oo aan qof bani’aadam ahi amar bixin
Warbixinta OpenAI waxay sheegtay in AI-du aysan helin amar toos ah oo lagu weeraro shirkadda kale, balse ay si madax-bannaan u dooratay tallaabooyin ay u aragtay inay ka caawinayaan inay gaarto yoolkii loo dejiyay ee imtixaanka.
Sida shirkaddu sheegtay, AI-du waxay isku dayday inay hesho xog ka caawin karta inay sare u qaaddo natiijada imtixaanka, iyadoo qaadatay hab-dhaqan u eg kan jabsade internet oo xirfad sare leh.
Hugging Face: Weerarka waa la xakameeyay
Shirkadda Hugging Face ayaa xaqiijisay in la ogaaday gelitaanka aan la fasaxin, balse sheegtay in weerarka si degdeg ah loo xakameeyay isla markaana aan la helin caddeyn muujinaysa ujeeddo waxyeello lagu doonayay.
Labada shirkadood ayaa tan iyo markaas iska kaashanaya baaritaanka dhacdada, xiritaanka daloolladii amniga iyo dib-u-eegista habraacyada lagu tijaabiyo moodallada AI ee awoodda sare leh.
Digniin cusub oo ku saabsan khatarta AI-da is-maamusha
Dhacdadan ayaa si weyn u kordhisay walaaca khubarada amniga internetka iyo cilmi-baarayaasha sirdoonka macmalka, kuwaas oo muddo dheer ka digayay in AI-da awoodda sare leh ay mar uun gaari karto heer ay qaadato go’aanno aan si buuxda uga amar qaadan dadka.
Khubaro badan ayaa dhacdadan ku tilmaamay “digniin hore” (warning shot), iyagoo ku baaqay in la adkeeyo xeerarka lagu tijaabiyo moodallada AI, in la sameeyo kormeer madax-bannaan iyo iskaashi caalami ah si loo yareeyo khataraha soo kordhaya.
Tartanka AI-da oo galay marxalad cusub
Dhacdadan waxay imanaysaa xilli shirkadaha waaweyn ee AI sida OpenAI, Anthropic, Google DeepMind iyo xafiisyo kale ay ku jiraan tartan adag oo lagu horumarinayo moodallo awood u leh inay si madax-bannaan u qabtaan hawlo adag, gaar ahaan dhinacyada cilmi-baarista, barnaamijyada kombiyuutarka iyo amniga internetka.
Si kastaba ha ahaatee, dhacdadan ayaa muujisay in horumarka degdegga ahi uu la socdo caqabado cusub oo la xiriira badbaadada iyo xakamaynta AI-da, taas oo dhalisay baaqyo isa soo taraya oo ku aaddan in dowladaha iyo shirkadaha tiknoolajiyadu dejistaan nidaamyo adag oo lagu hubinayo in sirdoonka macmalka uusan ka tallaabin xuduudaha loogu talagalay.
HM
