I-OpenAI isemgqeni wokwenza inguquko enkulu kunazo zonke kusukela kwethulwa i-ChatGPT. Inkampani ihlela ukuguqula inkundla yayo isuke esetshenziswa njengohlelo lokuxoxa ibe isikhungo esiphelele sokukhiqiza nezinsiza ezenzakalelayo. Lolu hlelo luza ngesikhathi lapho imboni yobuhlakani bokwenziwa ihlangabezana nezinguquko ezinkulu, izinhlobo zolimi oluphambili sezishintshe indlela abantu abasebenzisana ngayo namakhompyutha. Isinyathelo esilandelayo akusikho nje ukuphendula imibuzo, kodwa ukwenza izenzo eziphathekayo. Inhloso ukuthi i-AI ingabe iseyithuluzi lokubuza nje, kodwa ibe umsizi wezinto zonke onamandla okwenza imisebenzi eyinkimbinkimbi egameni lomsebenzisi.
Kusuka Engxoxweni Kuya Ezenzweni: Umngcele Omusha We-AI
Lapho i-ChatGPT yaqala ukwethulwa, ubuhle bayo obukhulu kwakuwukwazi ukuqonda ulimi lwemvelo nokukhiqiza izimpendulo ezinemininingwane emizuzwaneni, okwaholela ekutheni ithole izigidi zabasebenzisi futhi yandise ukusetshenziswa kobuhlakani bokwenziwa emhlabeni wonke. Kodwa isimo sishintshe ngokushesha. Namuhla, abasebenzisi bafuna okungaphezu kwezimpendulo: bafuna i-AI yenze imisebenzi, ithole ulwazi, ikhiqize amadokhumende, ihlele amakhalenda, yenze ukuhlaziya, futhi isebenze ngokuzenzakalelayo. Kulesi simo, umbono omusha we-OpenAI we-ChatGPT uvela. Inkampani ithuthukisa amandla okuhlanganisa imisebenzi eminingi ekuhlangenwe okukodwa, okuvumela umsebenzisi ukuthi achaze umgomo futhi ayeke izinsiza ezenzakalelayo zenze izinyathelo ezidingekayo. Isibonelo, uma ucele ukuhlela uhambo oluya eBrasília — ukuthola izindiza, ukubhuka indawo yokulala, ukwakha uhlelo lokuhamba, nokuvumelanisa ikhalenda — isizukulwane esisha sezinsiza ezenzakalelayo sizokwazi ukwenza ingxenye enkulu yale misebenzi ngokuzenzakalelayo, kunokuba sinikeze iziphakamiso nje.
Umqondo Wezinsiza Zobuhlakani Ezenzakalelayo
Igama elithi "insiza yobuhlakani" seliyinye lezinto ezibalulekile embonini yezobuchwepheshe ngonyaka we-2026. Ngokungafani nohlelo lokuxoxa oluvamile, insiza yobuhlakani inamandla okuhlela, ukwenza, nokuqapha imisebenzi. Lezi zinhlelo zingahlukanisa imigomo eyinkimbinkimbi ibe izinyathelo ezincane, zenze izenzo kuzinhlelo ezixhumene, futhi zibuyisele imiphumela kumsebenzisi. Empeleni, insiza yobuhlakani ingakwazi ukwenza lokhu okulandelayo:
- Ukuphenya ulwazi kwi-inthanethi;
- Ukuqhathanisa amanani nezinsiza;
- Ukuphatha amadokhumende;
- Ukukhiqiza imibiko;
- Ukubuyekeza amaspredishidi;
- Ukuthumela imiyalezo;
- Ukudala izethulo;
- Ukwenza imisebenzi yebhizinisi elandelanayo.
I-OpenAI ikholelwa ukuthi le modeli imelela ukuguquka kwemvelo kobuhlakani bokwenziwa besimanje. Amandla okwenza ngokuzenzakalelayo imisebenzi emhlabeni wedijithali akhuphula inani lenkundla ngaphezu kwengxoxo nje.
Umjaho Wohlelo Olukhulu Lwe-AI
Emuva kwesigcawu somboni, kuya ngokuya kukhula umbono wokuthi i-ChatGPT ingaguquka ibe yilokho okubizwa ngokuthi "uhlelo olukhulu," umqondo odume e-Asiya ngezinkundla ezihlanganisa izinsiza ezahlukene endaweni eyodwa. Ngokusebenzisa lo mbono emhlabeni wobuhlakani bokwenziwa, i-OpenAI ingaguqula i-ChatGPT ibe isikhungo sokusebenza sezinto zedijithali. Esikhundleni sokusebenzisa izinhlelo ezahlukene zokuxhumana, ukukhiqiza, ukuhlela, nokuphenya, abasebenzisi bazokwenza konke endaweni eyodwa. Le ndawo ehlanganisiwe ingase ihlanganise lokhu:
- Izinsiza zomuntu siqu;
- Amathuluzi okukhiqiza;
- Ukukhiqiza izithombe;
- Ukuhlaziya amadokhumende;
- Ukuhlela ngosizo;
- Ukuphenya okuthuthukile;
- Ukusebenza ngokuzenzakalelayo kwamabhizinisi;
- Izinsiza ezenzakalelayo ezikhethekile.
Kochwepheshe, lokhu kuhlangana kunamandla okuchaza kabusha indlela esisebenzisa ngayo amakhompyutha namaselula. Lolu shintsho lwenzeka phakathi komjaho omkhulu wezobuchwepheshe kusukela ekusakazweni kwe-inthanethi. Izinkampani ezinjenge-Google, Microsoft, Meta, ne-Anthropic zitshala izigidi ezigidini zamadola ezinhlelweni ezithuthukile zobuhlakani. I-Google inweba uhlelo lwayo i-Gemini; i-Anthropic ikhula nge-Claude; i-Meta itshala emamodelini avulekile; futhi i-Microsoft inweba ukuhlanganiswa kobuhlakani emikhiqizweni yamabhizinisi. Kulesi simo sokuncintisana, i-OpenAI ifuna ukugcina ubuholi bayo ngokuthembela ezinsizeni ezenzakalelayo ukuze inwebe inani lenkundla kubo bobabili abasebenzisi ngabanye namabhizinisi.
Iqhaza Le-Codex Nokuhlelwa Okusizwa
Enye ingxenye ebalulekile yamasu ihlanganisa ukuthuthuka kwamathuluzi okuhlela. I-OpenAI iyaqhubeka nokutshala emasistimu akwazi ukusiza abathuthukisi ezigabeni zonke zokuthuthukisa isofthiwe. Intuthuko ihlanganisa lokhu:
- Ukukhiqiza ikhodi ngokuzenzakalelayo;
- Ukulungisa amaphutha;
- Imibhalo ehlakaniphile;
- Ukuhlolwa okuzenzakalelayo;
- Ukuhlaziya ukuphepha;
- Ukuthuthukisa ukusebenza kahle.
Ucwaningo olwenziwa izikhungo ezinjenge-Yunivesithi yaseStanford ne-MIT lukhombisa ukuthi amathuluzi obuhlakani angakhuphula kakhulu umkhiqizo wabathuthukisi uma esetshenziswa ngendlela efanele. Lokhu kuthuthuka kubeka i-OpenAI emakethe yokuthuthukisa isofthiwe, kunikeza umbono ohambisana nohlelo olukhulu.
Amabhizinisi Nezinselelo Zokuzimela
Nakuba i-ChatGPT ithole ukuthandwa okukhulu phakathi kwabathengi, imakethe yamabhizinisi isiyithuba elikhulu kakhulu le-OpenAI. Izinkampani ezinhlobonhlobo zifuna ukusebenza ngokuzenzakalelayo, ukunciphisa izindleko zokusebenza, nokukhuphula umkhiqizo. Kulesi simo, izinsiza ezenzakalelayo zingathatha imisebenzi eyayiyenziwa amathimba aphelele. Phakathi kokusetshenziswa okuthembisayo kwamabhizinisi kukhona:
- Ukunakekela amakhasimende;
- Ukuhlaziya idatha;
- Ukukhiqiza imibiko;
- Ukuqapha ukusebenza;
- Ukuphatha amadokhumende;
- Ukusekela ezobuchwepheshe ngaphakathi;
- Ukusebenza ngokuzenzakalelayo izinqubo eziphindaphindwayo.
Ukusetshenziswa kwalezi zinhlelo kungakhiqiza imali eyongwayo ebalulekile ezinhlanganweni zazo zonke izinhlobo. Kodwa-ke, uma ubuhlakani buthola amandla amaningi, izinselelo eziphathelene nokuphepha ziyakhula. Ochwepheshe baxwayisa ukuthi izinsiza ezenzakalelayo ezingenza izenzo eziphathekayo zidinga izindlela eziqinile zokuqapha. Izinto ezikhathaza kakhulu zihlanganisa:
- Ukuvikela ulwazi lomuntu siqu;
- Ukulawula izimvume;
- Ukuvimbela ukukhwabanisa;
- Ukubeka izinqumo obala;
- Ukuqapha imisebenzi;
- Ukuvimbela ukusetshenziswa kabi.
Ngenxa yalokho, ohulumeni nezinhlangano zamazwe ngamazwe basebenza ekwenzeni imithetho ethile yokusetshenziswa kobuhlakani ngomthwalo. Izinhlangano ezinjenge-NIST ne-IYunithi Yezwe YaseYurophu zithuthukisa izinhlaka zokulawula ezigxile ekuphathweni kobuhlakani.
Kumsebenzisi ojwayelekile, ukuthuthuka kwe-ChatGPT kungamela ushintsho olukhulu njengokuvela kwamaselula. Esikhundleni sokushintsha phakathi kwezinhlelo, abantu bangathuma imisebenzi ngqo ezinsizeni ezenzakalelayo, njengokuhlela izimali, ukuhlela uhambo, ukuphatha izifundo, nokulawula izibopho. Ngesikhathi esifanayo, kuba nesidingo sokufunda izindlela zokusebenza zedijithali ezigxile ekusetshenzisweni kobuhlakani ngokuqapha; abasebenzisi kudingeka baqonde imikhawulo yalezi zinhlelo, baqinisekise ulwazi olubalulekile, futhi baqaphe izinqumo ezenziwa ngokuzenzakalelayo. Nakuba i-OpenAI ingakayidaluli yonke imininingwane yamasu ayo esikhathini esizayo, izimpawu zikhombisa ukuthi inkampani ithembele kakhulu ekuhlanganiseni izinsiza ezenzakalelayo nokusebenza okuthuthukile. Okulindelwe emakethe ukuthi izici ezintsha zizokwethulwa kancane kancane, okuvumela i-ChatGPT ukuthi iguquke isuke kumsizi wengxoxo iye enkundleni esebenzayo ephelele. Uma lo mbono ufezeka, indlela esisebenzisana ngayo namakhompyutha ingashintsha kakhulu, esikhundleni sezinhlelo ezihlukanisiwe kube ingqalasizinda lapho abasebenzisi bechaza izinhloso nezinsiza zobuhlakani zenze ingxenye enkulu yomsebenzi.