Esikhathini esedlule asizange sikhulume ngakho I-OpenAI kanye nehansi lakhe elizalela amaqanda egolide, ChatGPT. Futhi ukuthi izindaba zakamuva azihlangene kangako nokusebenza kwayo futhi zihlobene kakhulu nethuba lokuvimba i-chatbot edumile. Isizathu, lobo bumfihlo abunakuqinisekiswa, ukuthi inkampani ingakwazi konke ngathi ngoba sibatshela noma yini. Kuleli sonto, le nkampani ibonakala ithathe igxathu lokugwema izinkinga kwamanye amazwe.
Le yindlela i-OpenAI ekwenze kwaziwa ngayo esihlokweni eshicilelwe ngo-April 25, ngaphansi kwamahora angu-24 edlule. Isihloko sayo sithi "Izindlela ezintsha zokuphatha idatha yakho ku-ChatGPT", futhi okuqukethwe ngokuyisisekelo kuyincazelo yenketho entsha ongayisebenzisa. singenza umlando wengxoxo ungasebenzi. Singafinyelela kuyo ngokukhetha umsebenzisi wethu (ngezansi kwesokunxele), ngokuchofoza "Izilungiselelo", kokuthi "Izilawuli zedatha" ngokuchofoza okuthi "Bonisa" bese uvala iswishi.
I-OpenAI ingase ibe nezinkinga uma ingakuqinisekisi ubumfihlo
Umbhalo ochazayo ngaphansi kwaleyo shintshi uthi izingxoxo ezintsha [uma zihloliwe) zigcinwa emlandweni wethu futhi sivumela ukuthi zisetshenziswe ukuqeqesha i-ChatGPT. Izingxoxo ezingalondoloziwe zizosuswa ezinhlelweni ze-OpenAI ezinsukwini ezingama-30. Noma kunjalo, izingxoxo zizolondolozwa kuleyo nyanga, kodwa, bathi, ukuzibuyekeza kuphela lapho kudingekile ukuze kuqinisekiswe ukuthi akukho ukuhlukumeza.
Esinye isivikelo sabasebenzisi ukuthi, uma sisebenzisa i-ChatGPT ngohlelo noma isevisi esebenzisa eyakho Business APIs, idatha yomsebenzisi ngeke isetshenziselwe ukuqeqesha i-chatbot ngokuzenzakalela. Lokhu kuzoza ezinyangeni ezimbalwa ezizayo.
Into yokugcina abasitshele ngayo kuleli nothi yi kungenzeka ukuthumela idatha yethu yokusetshenziswa ku-ChatGPT futhi uqonde ukuthi yiluphi ulwazi olugcina i-ChatGPT. Uma sinquma ukuzithekelisa, sizothola ulwazi ku-imeyili yethu yokubhalisa. Isikhathi sokuthola leyo imeyili sizoncika ezintweni ezingachazwanga, kodwa ngiyicelile futhi, ngemva kwemizuzu embalwa yokulinda, angikakatholi lutho.
Kungase kube ngenxa yezinye izizathu, kodwa kubonakala kimi ukuthi lokhu kunyakaza kuyimpendulo ezinsongweni zokuvinjelwa eziwiswe ngamazwe afana ne-Italy, iJalimane neSpain, okwakubheka ukubuza ukuthi omakhelwane bethu baseYurophu benzani.
Ithiphu elilodwa: ukuphathwa komlando okwenziwa ngesandla
Ngingaluleka ngenye indlela. Njengomsebenzisi obesebenzisa i-OpenAI chatbot ukubuza imibuzo eminingi, ngisho nokukhuluma ngamagama ezingoma noma izincazelo zamamuvi, umlando wami ubugcwele izingxoxo ezingenakubalwa. Iqiniso liwukuthi ungambuza ukuthi "Ubani igama lomculi we-Ad Infinitum" futhi lokho kugcinwa njengengxoxo, ngisho noma bekuwumbuzo kuphela nempendulo yokonga isikhathi. Ngisanda kunikela susa wonke umlando futhi ngiwuqale phansi.
Inketho eyodwa iwukuba ungalondolozi lutho, kodwa uma sihlanganise okuthile esikuthandayo, ngeke siphinde sikubuyekeze lokho. Ngakho-ke, umbono omuhle, noma okubonakala kimina, uwukwenza lokho kanye engikwenzayo manje: Ngiqala izingxoxo ezivamile, futhi lapho ukungabaza kwami kusuliwe noma ngithola engikufunayo, Ngiyisusa mathupha. Futhi angikwenzeli nje ubumfihlo, kodwa ukwenza yonke into ihleleke kakhudlwana. Impela, uma ngizwa sengathi ngibuza into exakile noma yezokwelapha (yize kungcono ukuya kuchwepheshe, yinto esesivele siyenze nge-Google), ngisusa leyo ngxoxo ukuze ibe yimfihlo.
Noma ngabe yikuphi ukusetshenziswa okwenziwayo, lokhu kungenzeka sekuvele kukhona, futhi manje kukumazwe ukuthi aqinisekise ukuthi abakushoyo kuyiqiniso yini futhi kwanele.