Riaan I-Apple isiphenduke inkampani yakamuva yokuvimbela ukusetshenziswa kwangaphakathi kwe-ChatGPT nemikhiqizo efanayo, okuxakayo njengoba nje ne-OpenAI chatbot iza ku-iOS ngohlobo lohlelo lokusebenza lweselula.
Izindaba zalesi senzo zivezwe izolo ngu I-Wall Street Journal, ebuyekeze idokhumenti ye-Apple yangaphakathi eyazisa abasebenzi ngokuvinjelwa. Ngokwalo mbhalo, ukukhathazeka kwe-Apple kuhambisana nezinye izinhlaka eziye zavimbela i-ChatGPT ukuthi isetshenziswe ngaphakathi, okungukuthi i-AI ingachitha imininingwane yangaphakathi ebucayi okwabelwana ngayo nayo.
i-apula kubikwa ukuthi ivimbele ithuluzi lekhodi le-GitHub elizenzakalelayo, Umshayeli wendiza, kanjalo. Amahemuhemu abelokhu ezulazula mayelana nezinhlelo ze-Apple ze-AI isikhathi eside, kanti inkampani kungenzeka isebenze yodwa i-LLM ukuze izimbangi i-ChatGPT ne-Google Bard.
UCupertino akayedwana esinqumweni sakhe sokwenqabela ukusetshenziswa kwe-chatbot esekelwa yi-Microsoft: Ijoyina i-chatbot ekhula njalo. uhlu yezinkampani ezifana ne-Amazon kanye nenani lamabhange ahlanganisa i-JPMorgan Chase, i-Bank of America, i-Citigroup, i-Deutsche Bank nokunye okunjalo.
Imbangi yakwa-Apple i-Samsung nayo ithuthele ekuvimbeleni i-ChatGPT ukuthi isetshenziswe ngaphakathi – kabili – ngenxa yamaphutha. AbakwaSamsung basusa ukuvinjelwa ukusetshenziswa kwabasebenzi kwe-ChatGPT ngoMashi, kodwa esikhathini esingaphansi kwenyanga kwaba njalo kwembulwa ngabezindaba baseKorea abasebenzi bakwaSamsung bacele i-ChatGPT usizo lokuxazulula iziphazamisi zekhodi yomthombo, ukulungisa isofthiwe esetshenziselwa ukuqoqa ukulinganisa nokukhipha idatha kanye nokuphendula amanothi omhlangano abe amaminithi.
I-Samsung kubekwe kabusha ukuvinjelwa kwayo kwe-ChatGPT ekuqaleni kwale nyanga ukuvimbela izigameko ezifanayo ukuthi zingaphinde zenzeke.
Inkinga nge-ChatGPT, i-Google Bard kanye ne-LLM bots ukuthi idatha efakwe kuzo ivame ukusetshenziselwa ukuqeqesha ama-bots, i-ejensi yezinhloli yase-UK, i-GCHQ, eye yaxwayisa kalula ukuthi ingaholela kalula kulwazi oluyimfihlo lwebhizinisi. ukuphinda uma abanye bebuza imibuzo efanayo.
Imibuzo iyabonakala nakubahlinzeki be-bot, abafana ne-OpenAI ne-Google, abangabuyekeza ngokwabo okuqukethwe okuphakelwa amamodeli olimi lwabo, baphinde babeke engcupheni yokuvezwa kwezimfihlo zebhizinisi eziqashwe eduze.
Izingozi ziyenzeka futhi
Kanye nobungozi bokuthi i-bot yabelana ngolwazi oluyimfihlo lapho izama ukusiza abanye, kunethuba lokuthi izinkampani ezifana ne-OpenAI zimane nje aziyifaki ikhodi isofthiwe ehamba phambili.
NgoMashi, i-OpenAI kuvume ukuthi iphutha kulabhulali yomthombo ovulekile redis-py kubangele izingcezu ze-tête-à-têtes zabantu nge-ChatGPT ukuthi zibonwe abanye abasebenzisi. Leyo bug, umhlaziyi wedatha oholayo e-Kaspersky Vlad Tushkanov usitshele, kufanele kube isikhumbuzo sokuthi i-LLM chat bots ayinikezi ubumfihlo bangempela kubasebenzisi.
“I-ChatGPT ixwayisa ngokungena ngemvume ukuthi ‘izingxoxo zingabuyekezwa abaqeqeshi bethu be-AI’ … Ngakho-ke kusukela ekuqaleni abasebenzisi bekufanele bangalindeli neze ubumfihlo lapho besebenzisa idemo yewebhu ye-ChatGPT,” kusho uTushkanov.
I-OpenAI ngenyanga edlule kwengezwe ikhono labasebenzisi be-ChatGPT lokukhubaza umlando wengxoxo, ongagcini nje ngokufihla ingxoxo kubha eseceleni kusixhumi esibonakalayo se-ChatGPT, kodwa futhi ovimbela izingxoxo ezikhutshazwe umlando ukuthi zisetshenziselwe ukuqeqesha amamodeli we-OpenAI.
I-OpenAI ithe isazogcina izingxoxo izinsuku ezingama-30 uma umlando ukhutshaziwe, futhi izoba nekhono lokuzibuyekeza “uma kudingeka ukuthi iqaphe ukuhlukumeza, ngaphambi kokuyisusa unomphela,” kusho inkampani esekelwa yiMicrosoft.
Esimemezelweni esifanayo, i-OpenAI iphinde yathi maduzane izokhipha inguqulo yebhizinisi ye-ChatGPT enikeza amabhizinisi amandla okulawula ukusetshenziswa kwedatha yawo, lapho i-OpenAI yathi isho ukuthi izingxoxo ze-ChatGPT Business ngeke zisetshenziselwe ukuqeqesha ama-LLM ayo.
Sibuze i-OpenAI eminye imibuzo eyengeziwe mayelana Nebhizinisi le-ChatGPT, njengokuthi ingabe abasebenzi be-OpenAI basazokwazi yini ukubuka izingxoxo nokuthi zingakhishwa nini, futhi sizobuyekeza le ndaba uma sibuzwa. ®
