Umncintiswano wokuhola ubuhlakani bokwenziwa kokukhiqiza Ukungezwani kukhuphuke kakhulu emasontweni amuva nje. Ngemva Ukwethulwa kweGemini 3 I-OpenAI ye-Google inqume ukuhamba ngokushesha futhi iqhubekele phambili ukufika I-GPT-5.2, ukuphindaphinda okusha kwemodeli yayo eyinhloko, ihlose ukuqinisa ukusebenza kwe-ChatGPT emisebenzini eyinkimbinkimbi, ukuthuthukisa ukuzinza, nokunciphisa amaphutha ekusetshenzisweni kwansuku zonke.
Lokhu kwethulwa akuhloselwe ukuba yigxathu elikhulu, kodwa njengesibuyekezo esibalulekile ochungechungeni lwe-5. Kodwa-ke, inhlanganisela ye- ukusheshiswa kokuthunyelwa, izinguquko kumephu yomgwaqo yangaphakathi kanye nokugxila kakhulu ekucabangeni nasekukhiqizeni emsebenzini kubeka i-GPT-5.2 enhliziyweni yesu le-OpenAI lokugwema ukulahlekelwa yi-Google, i-Anthropic kanye I-DeepSeek v3.2, abanye abadlali abazibeke phezulu ezikhundleni zobuchwepheshe.
Ikhodi ebomvu ku-OpenAI kanye nokukhishwa kwangaphambi kwesikhathi: GPT-5.2
Isinqumo sokuletha i-GPT-5.2 sihlelwe ngaphakathi komongo we- ingcindezi enkulu yokuncintisanaUkwamukelwa okuhle kwe-Gemini 3—ikakhulukazi ekuhlolweni okuthuthukisiwe kokucabanga kanye nokubhala amakhodi—kwashukumisela i-OpenAI ukuthi isebenze ngaphakathi “ikhodi ebomvu.” USam Altman, i-CEO yenkampani, uthumele imemo ecela ukuthi izinsiza zigxile ekuthuthukiseni i-ChatGPT kanye nezinhlelo zesibili, njengezivivinyo ezithile zokwenza imali kanye nezici ezingabalulekile kakhulu epulatifomu, zimiswe.
Ngokusho kokuvuza okuhlukahlukene, isibuyekezo ekuqaleni sasihlelelwe ukuphela kukaDisemba, kodwa kubikwa ukuthi abaphathi banqume ukuletha ukuqaliswa kwayo emavikini ambalwa ukuvala igebe lokusebenza kanye nombono womphakathi elidalwe amamodeli wakamuva e-Google. Nakuba usuku oluqondile luhlala luncike ekulungisweni kobuchwepheshe okungenzeka ngomzuzu wokugcina, imithombo iyavuma ukuthi ishejuli yangaphakathi icindezelwe ukuze i-GPT-5.2 ifinyelele abasebenzisi nabathuthukisi abakhokhayo ngokushesha okukhulu.
Lokhu kuguquka kwemicimbi kufana, yize kungesilinganiso esihlukile, nesiqephu sango-2022, lapho ukwethulwa kwe-ChatGPT kwaphoqa i-Google ukuthi isheshise imephu yayo yomkhiqizo. Manje izindima sezishintshiwe, futhi yi-OpenAI ezama... qinisekisa kabusha isikhundla sayo sokulinganisa emakethe lapho amazinga okusebenza kanye namathuluzi okuqhathanisa amamodeli eshintsha cishe nsuku zonke.
I-GPT-5.2, ukuvela ngaphakathi kochungechunge lwezi-5 okugxile emsebenzini wolwazi
I-GPT-5.2 yethulwa njengokuqhubeka ngqo kwe-GPT-5.1, hhayi isizukulwane esisha ngokuphelele. Noma kunjalo, inkampani igcizelela ukuthi lesi sibuyekezo simelela intuthuko enkulu kulokho okubizwa ngokuthi umsebenzi wolwazi: ukuhlela, ukuhlaziywa kwemibhalo, ukumodela kwezezimali, ucwaningo lwesayensi, noma ukulungiselela imibiko eyinkimbinkimbi.
I-OpenAI ithi imodeli ilawula izimo ezindeKunciphisa amaphutha okucabanga futhi kuthuthukisa ikhono lokuhlanganisa ukulandelana kwezenzo namathuluzi angaphandle. Lokhu kuhlanganiswa kuyisihluthulelo semisebenzi edlula ukuphendula umbuzo olula, njengemiklamo enezinyathelo eziningi, ukubuyekezwa kwamadokhumenti okubanzi, noma ukuzenzekela okuyingxenye kwemisebenzi yebhizinisi.
Empeleni, i-GPT-5.2 ithembisa intuthuko ekudalweni kwe amaspredishithi anemininingwane, izethulo ezihlelekile, imidwebo yokusebenza, kanye nemibhalo yobuchwepheshe, ngenhloso yokuvumela izinkampani ukuthi zinikezele umsebenzi "owenziwe ngezandla" kumodeli ngaphandle kokuchitha isikhathi esiningi zilungisa futhi zibhala kabusha.
Izinhlobo ezintathu: Okusheshayo, Okucabangayo, kanye Nobuchwepheshe
Umndeni omusha we-GPT-5.2 uhlelwe ngendlela izingqimba ezintathu ezihlukene zokusetshenziswa, ngenhloso yokulungisa imodeli ukuze ivumelane nezidingo ezahlukene kanye namazinga ezindleko:
- I-GPT-5.2 InstantLe nguqulo ibeka phambili isivinini futhi yenzelwe imibuzo yansuku zonke, ukubhala okuvamile, ukuhumusha, ukubuyisa ulwazi, kanye nemisebenzi lapho isikhathi sokuphendula sibaluleke kakhulu kunokucabanga okujulile. Lokhu kuhlukahluka kuphinde kuzuze ezincazelweni ezizinzile kanye namaphutha ambalwa uma kuqhathaniswa nezinguqulo zangaphambilini.
- GPT-5.2 Ukucabanga: ingabe inguqulo ihloselwe ukucabanga okunezinyathelo eziningi kanye nokuphathwa kwemibhalo ebanziIgxile ekuhleleni izinhlelo okuyinkimbinkimbi, ukuhlaziywa kwedatha, imisebenzi yezibalo ethuthukisiwe, ukumodela kwezezimali, ukubuyekezwa kwenkontileka, kanye nokuhlela amaphrojekthi esikhathi eside. Yilapho i-OpenAI igxila khona kakhulu ekuthuthukisweni kwayo kokungaguquguquki kanye nokusetshenziswa kwamathuluzi ahlanganisiwe.
- I-GPT-5.2 ProIbekwe endaweni ephezulu yokusetshenziswa okudinga kakhulu, igxile ekuqondeni okuphezulu kakhulu ngaphakathi kwemikhawulo yezobuchwepheshe yamanje. Yimodeli ehloselwe labo ababeka phambili ikhwalithi yokucabanga kunesikhathi sokulibaziseka futhi abazimisele ukwamukela izindleko zokubala eziphakeme, njengamaqembu e-R&D, amahhovisi akhethekile, noma amaphrojekthi esayensi ayinkimbinkimbi.
Lokhu kuhlukaniswa kwezigaba kuhlose okungaphezu nje kokunikeza "imodeli enamandla kakhulu": kuhlose ukwenza ikhathalogi ivumelane namaphrofayili abasebenzisi ahlukene, kusukela kubasebenzisi abafuna izimpendulo ezisheshayo ku-ChatGPT kuya Izinkampani zaseYurophu abasebenzisa ama-ejenti angaphakathi kudatha yabo nge-API.
Ukusebenza kuma-benchmark e-GPT-5.2: ukucabanga, ikhodi, kanye nesayensi
I-OpenAI ihambisana nokuqaliswa ngebhethri ledatha elibeka I-GPT-5.2 ingaphezu kwe-GPT-5.1 Cishe kuzo zonke izigaba ezikhethe ukuzishicilela. Ekuhlolweni okufana ne-GDPval, eqhathanisa imiphumela yemodeli nochwepheshe babantu emisebenzini engu-44, i-GPT-5.2 ifinyelela ukunqoba noma ukulingana cishe ku-70,9% wamacala, kanye nokuthuthuka okukhulu emisebenzini ehilela ukudalwa kwezethulo, amadokhumenti okusebenza, kanye nezinto zezimali.
Ukuhlolwa okukhethekile njenge-GPQA Diamond—okugxile emibuzweni yezinga leziqu ku-physics, chemistry, kanye ne-biology—, I-GPT-5.2 Pro ifinyelela cishe ukunemba okungu-93%Lokhu kulandelwa eduze yi-Thinking variant, ewela ngaphansi kancane kodwa futhi ijikeleza kulowo mkhawulo. Ezibalweni ezithuthukisiwe, amaphuzu emodeli ku-FrontierMath (Tier 1-3) akhuphuka afinyelele ngaphezu kuka-40%, okuyisibalo esisekude kakhulu nokuphelela kodwa esiphakamisa intuthuko eqhubekayo ekhonweni lokulandela izintambo ezinengqondo ezinde nezihlelekile.
Isigaba sokubhala amakhodi siphinde sibhekane nokukhula. Ku-SWE-Bench Pro, ehlola ukuxazululwa kwezigameko zangempela ezindaweni zokugcina isofthiwe Futhi ngokunciphisa ingozi yokuthi imodeli ibone idatha ngaphambilini, i-GPT-5.2 Thinking ithuthukisa kuneyandulelayo ngamaphuzu amaningana, ifinyelela izinga lokuxazulula izinkinga elicishe libe ngu-55,6%. Emisebenzini eqinisekisiwe, lesi sibalo sikhuphukela cishe ku-80%, okuhunyushwa ngokusebenzayo njengokungenelela okuncane ngesandla kokubukeza ama-patches, ukulungisa kabusha, kanye nezingxenye eziphelele.
Ukwenza ukuhlolwa okwengeziwe kobuchwepheshe, njenge-ARC-AGI (ukucabanga okungaqondakali kanye nokutholwa kwephethini) noma amasethi athile esayensi kanye nokuhlela, imodeli ingaphezulu kwe-GPT-5.1 futhi, ngokusho kwamagrafu ashicilelwe yi-OpenAI, ingaphambi kwe-Gemini 3. I-Grok 4 Ngokushesha kanye noClaude Opus 4.5 ezivivinyweni eziningana eziyinkimbinkimbi zokucabanga. Lezi zinhlobo zezilinganiso, nakuba zihlala ziphikisana ngokumelela kwazo, zingenye yezimpikiswano eziyinhloko inkampani ezama ngazo. Kholisa abatshalizimali kanye namakhasimende amakhulu ukuthi ubuholi bezobuchwepheshe bezimbangi zabo, okungenani, bungaphikiswana ngakho.
Umthelela emisebenzini yangempela: ezezimali, ukuhlaziywa kwamadokhumenti, kanye nama-ejenti
Ngaphandle kwezinombolo, i-OpenAI igcizelela ukuthi ukuthuthuka kuyabonakala ku imisebenzi yansuku zonkeEkulingiseni kwangaphakathi okulingisa imisebenzi yabahlaziyi bezezimali—njengokwakha amamodeli ezifundazwe ezintathu noma imisebenzi yokuthenga eqondiswe nge-levered—Ukucabanga kwe-GPT-5.2 ngabe kusuke kumaphuzu aphakathi aseduze no-59% kuya kwelinye elingaphezu kuka-68%, kunciphisa amaphutha okubala kanye nesidingo sokulungiswa okulandelayo.
Izinkampani ezifana neNotion, Box, Shopify, Harvey, kanye neTriple Whale, ebezivele zisebenzisa amamodeli angaphambilini enkampanini, kubikwa ukuthi zibone intuthuko ebhizinisini. ukuzinza kwama-ejenti asekelwe kumathuluziLokhu kuphumela ekuxhumaneni okungcono phakathi kwezingcingo eziningi ze-API, izinyathelo eziphakathi ezihambisanayo, kanye nokuvinjelwa okumbalwa kokugeleza okude. Kwezinye izimo, ngokusho kwalokhu bufakazi, kuye kwenzeka ukuthi kufakwe esikhundleni izakhiwo zama-multi-agent ezibuthakathaka nge-ejenti eyodwa esekelwa yi-GPT-5.2, enamathuluzi axhunyiwe angaphezu kwamashumi amabili kanye nesidingo esincane sokuqapha okuqhubekayo.
Kumaqembu omkhiqizo, ukwesekwa, kanye nokuthuthukiswa ngaphakathi kwezinhlangano zaseYurophu, lolu hlobo lwezinguquko luhumusha ithuba lokwakha abasizi bangaphakathi abazokwakha Bacubungula izinkontileka ezinde, imibiko yemithetho, noma imibhalo yobuchwepheshe. ngaphandle kokulahlekelwa yingxoxo ngemuva kwamakhasi amaningi noma amafayela amaningi ahlobene, into ebaluleke kakhulu emikhakheni elawulwayo njengezezimali, ezempilo noma amandla.
Ukubuka, ukuxhumeka kwezithombe kanye nokuqonda amadokhumenti amade ku-GPT-5.2
Ingxenye ye-multimodal nayo ithatha isinyathelo esiya phambili. Ekuhlolweni okufana ne-CharXiv Reasoning—egxile ezibalweni zesayensi—i-GPT-5.2 ihlukanisa amaphutha okuhumusha uma kuqhathaniswa ne-GPT-5.1. Ku-ScreenSpot-Pro, ukuhlolwa okulinganisa ikhono lokuqonda izixhumi ezibonakalayo eziyinkimbinkimbiImodeli ikhulisa ukunemba kwayo kumanani asondele ku-86%, okuwusizo kakhulu kumaphaneli okulawula okufunda, amadeshibhodi noma imidwebo yesofthiwe.
Ngokuphathelene nenkumbulo yomongo, i-GPT-5.2 isondela ekusebenzeni okuphelele kuzinhlobo ze-MRCRv2 kumakhulu ezinkulungwane zamathokheni. Ngokwezwi elingokoqobo, lokhu kusho ukuthi ingaphatha imiqulu emikhulu yombhalo —ukubonisana ngemibiko, amafayela, ukuhlolwa kwezobuchwepheshe noma imibhalo yezemfundo—ukugcina izinkomba zangaphakathi kanye nokuvumelana phakathi kwezigaba, into izinhlangano eziningi zaseYurophu eziyibona njengesimo esibalulekile sokuphathisa izinqubo ezibucayi kumodeli ye-AI.
Lokhu kuhlanganiswa kombono othuthukisiwe kanye nekhono elikhulu lokuqukethwe kuvula ithuba lokusetshenziswa okunezinhloso ezinkulu, njengokuthi isibuyekezo esihlanganyelwe sezethulo, amaspredishithi, kanye namadokhumenti e-PDF ngaphakathi kokuhamba okufanayo, noma ukuhlolwa kwezixhumi zewebhu namathuluzi angaphakathi ukuze kube lula ukwesekwa kobuchwepheshe kanye nokuhlaziywa kokusebenziseka.
Amaphutha ambalwa, kodwa kunesidingo sokuphathwa ngabantu
Esinye sezithembiso eziphindaphindwa kakhulu yinkampani yi- ukunciphisa amaphutha ezimpendulweniI-OpenAI ithi i-GPT-5.2 Thinking ikhiqiza cishe izimpendulo ezinephutha ezingaphansi kuka-30% kune-GPT-5.1, nokuthi jikelele izinga lezimpendulo ezinephutha liyehla lisuka cishe ku-8,8% liye cishe ku-6,2%.
Noma kunjalo, inkampani igcizelela ukuthi imodeli isalokhu iyinto engenzeka nokuthi isitatimende esisodwa esingalungile singadinga ukubuyekezwa ngesandla kwayo yonke imiphumela, ikakhulukazi ezimweni ezibucayi noma ezilawulwayo. Yingakho igcizelela ukuthi i-GPT-5.2 kufanele ibhekwe njenge- ithuluzi lokusekela ukucabangahhayi esikhundleni sokwahlulela kwabantu, ikakhulukazi ezindaweni ezifana nezempilo, ezezimali, umthetho, noma ucwaningo lwezemfundo.
Ezindaweni ezibucayi—ngokwesibonelo, izingxoxo mayelana nempilo yengqondo noma ukucindezeleka ngokomzwelo—inkampani ithi ithuthukise izilawuli zokunciphisa izimpendulo ezingafaneleKodwa-ke, uyavuma ukuthi kusenendawo yokuthuthukisa. Lokhu kucatshangelwa kubaluleke kakhulu eYurophu, lapho uhlaka olusha lokulawula lwe-AI lunezela khona izibopho ezengeziwe maqondana nokucaca, ukuphepha, kanye nokuphathwa kwezingozi.
Umnikelo emsebenzini wesayensi nowezibalo
I-OpenAI iphinde iveze i-GPT-5.2 njengethuluzi elenzelwe ukukhuthaza intuthuko yesayensiLe nkampani ithi i-Series 5 isivele inezinhlelo zokusebenza kwizibalo, i-physics, i-biology, isayensi yamakhompyutha, isayensi yezinkanyezi, kanye nesayensi yezinto zokwakha, nokuthi ngenguqulo entsha la macala aba njalo.
Ku-GPQA Diamond, enye yezindlela zokulinganisa zokuhlola ukuqonda okuthuthukile kwesayensi, I-GPT-5.2 Pro kanye nokucabanga kudlula ukunemba okungu-92%Lo mphumela uhunyushwa yinkampani njengophawu lokuthi imodeli ingasiza abacwaningi ukuthi bahlole imibono, babuyekeze izincwadi, noma babeke ubufakazi. Kwesinye isimo esibhaliwe, kubikwa ukuthi i-GPT-5.2 Pro inegalelo ekuxazululeni inkinga evulekile kumbono wokufunda wezibalo, yize lokhu kuhlala kuncike ekuqinisekisweni komuntu okulandelayo.
Nokho, i-OpenAI ngokwayo icacisa ukuthi lezi zinhlelo kufanele ziqondwe njenge abasizi besigaba sokuhlola yomsebenzi wesayensi: kuyasiza ekwakheni izibikezelo, ukuvuselela imibono noma ukusikisela izinyathelo eziphakathi, kodwa ngaphandle kokususa indima ebalulekile yochwepheshe uma kukhulunywa ngokuqinisekisa imiphumela, ukuhumusha ubufakazi kanye nokufaka iziphetho ngokwezimo.
Ukufakwa ku-ChatGPT kanye nokufinyelela nge-API
I-GPT-5.2 iqala ukusetshenziswa ngezigaba ku I-ChatGPT yabasebenzisi abakhokhayoLokhu kufaka phakathi izinhlelo ze-Plus, Pro, Go, Business, kanye ne-Enterprise. Akubona bonke ababhalisile abazobona imodeli entsha ngesikhathi esisodwa, njengoba i-OpenAI ikhetha ukusebenzisa ukufinyelela ngezigaba ukuze igweme izinkinga zomthamo, okungabonakala eYurophu njengokuqaliswa kancane kancane ezinsukwini ezimbalwa.
Ezinyangeni ezintathu ezizayo, i-GPT-5.1 izohlala itholakala njengemodeli yefa ngaphakathi kwe-ChatGPT ngaphambi kokuthatha umhlalaphansi kokugcina, ukuze izinhlangano ezithembele emisebenzini yokusebenza emisiwe zikwazi hlela inguquko ngaphandle kokuphazamiseka okungazelelwe. Lokhu kuhlala ndawonye kwesikhashana kwenza kube lula ukuhlola i-GPT-5.2 ngesikhathi esifanayo nangokulungisa izixwayiso, izilawuli zangaphakathi, kanye nezinqubo zokuqinisekisa.
Ku-API, igama lesizinda ligcina ukuxhumana okuvamile: i-Instant variant ivela njenge okwakamuva kwengxoxo ye-gpt-5.2Inguqulo ethi Thinking ibizwa ngokuthi I-gpt-5.2 kanye ne-Pro njengoba i-gpt-5.2-proAbathuthukisi bangashintsha izinga lokucabanga kunketho ye-Pro, ngezinga elisha le-xhigh elenzelwe amaphrojekthi lapho ikhwalithi yechungechunge elinengqondo ibaluleke kakhulu kune-latency noma izindleko.
Amanani, ukusebenza kahle kwe-GPT-5.2, kanye nokugxila kumakhasimende ebhizinisi
Ngokwezomnotho, i-GPT-5.2 iza ne izimali eziphakeme ngamathokheni ayisigidi I-GPT-5.1 ye-OpenAI ibeka intengo eyisisekelo cishe ku-$1,75 ngesigidi samathokheni okufaka kanye namathokheni okukhipha angu-$14 ngesigidi, ngezaphulelo ezingu-90% zokufakwayo okugciniwe. Uhlobo lwe-Pro lukhulisa izindleko, njengoba izibalo zikhuphukela kumakhulu amaningana amathokheni okukhipha ngesigidi kumalungiselelo alo okucabanga adinga kakhulu.
Le nkampani ithi ukusebenza kahle kakhulu kwale modeli kuvumela ukwehla kwezindleko ezisebenzayo ngomsebenzi ngamunye, ikakhulukazi ezimweni lapho i-GPT-5.2 ithatha isikhathi esincane ukufika empendulweni evumelekile, idinga ukuzama kabusha okumbalwa, futhi yenza amaphutha ambalwa. amaphutha ambalwa adinga ukwenziwa kabusha komsebenziNoma kunjalo, isakhiwo samanani siklanyelwe ngokucacile ukusetshenziswa kwebhizinisi kanye nokuthuthukiswa okujulile, kunokuba kube ukuhlolwa okwenziwa kanye kuphela.
Ku-ChatGPT, i-Plus kanye nokubhaliselwe okuphezulu kugcina amanani abo ajwayelekile, kushintsha ingxenye enkulu yezindleko ezengeziwe ekusetshenzisweni kwe-API. Ezinkampanini eziningi zaseYurophu kakade ezihlanganisa i-ChatGPT kuma-intranet, amathuluzi okukhiqiza, noma abasizi bangaphakathi, lokhu kungasho ukuthi lungisa kabusha izabelomali futhi unqume ukuthi yiziphi izinqubo ezifanele ukuthuthela ku-GPT-5.2 nokuthi yiziphi ezingaqhubeka nokusebenza namamodeli angaphambilini, angabizi kakhulu.
Ingqalasizinda, ukuphepha kanye nengcindezi yokulawula
Ukufakwa kwe-GPT-5.2 kuncike, njengasezizukulwaneni ezidlule, engqalasizinda ye I-Microsoft Azure kanye ne-NVIDIA GPUs (kufaka phakathi imindeni ye-H100, H200, kanye ne-GB200-NVL72). I-OpenAI izibophezele ekutshalweni kwezimali okuyizigidi eziningi emandleni okusebenzisa amakhompyutha ukusekela lawa mamodeli asemngceleni, umdlalo othwala izingozi zezimali futhi odinga inkampani ukuthi ihlale ifuna imithombo emisha yemali engenayo. Iphinde ihlole amamodeli okulinganisa isisindo avulekile njenge I-GPT OSS.
Ngesikhathi esifanayo, inkampani yethula izinyathelo ezengeziwe maqondana ukuphepha nokuvikelwa kwezinganeEsinye sezinyathelo ezihlaba umxhwele kakhulu ukuqaliswa kohlelo olukwazi ukulinganisa iminyaka yabasebenzisi, ngenhloso yokuguqula izimpendulo ze-ChatGPT kulabo abangaphansi kweminyaka engu-18 nokuvula indlela "yemodi yabantu abadala" yesikhathi esizayo enezilawuli ezithuthukisiwe. Lezi zinhlobo zezindlela zihambisana nezidingo zomthetho eziya ngokuya ziqina kokubili e-European Union nase-United States.
I-OpenAI iyavuma ukuthi izinhlelo zayo ngezinye izikhathi zingaba zimbi kakhulu, okusho ukuthi zenqaba izicelo ezingaphuli imigomo, futhi ithi iyasebenza ukuze ukuphepha okungcono kanye nokusetshenziswa kwebhalansiLe nkampani iphinde igcizelele ukuthi noma yiziphi izinguquko ezifanele ekutholakaleni kwezinguqulo zangaphambilini—njenge-GPT-5.1, i-GPT-5, noma i-GPT-4.1 ku-API—zizomenyezelwa kusenesikhathi, uphawu lokuqhubeka kwamakhasimende asathembele kulawo mamodeli.
I-GPT-5.2 yethulwa njengesibuyekezo somjikelezo esizama ukuhlanganisa ukucabanga okuthuthukisiwe, isivinini, kanye nokuzinza nesu eligxile kakhulu ekusetshenzisweni kobungcweti kanye namabhizinisi. Uma ukuthuthukiswa kokubhala amakhodi, isayensi, ukuhlaziywa kwamadokhumenti, kanye nokuphathwa kwezimo ezibanzi kuhlanganiswa emisebenzini yansuku zonke, imodeli ingaba ithuluzi elifanele izinhlangano zaseYurophu ezifuna ukwenza ngokuzenzakalelayo ingxenye yezinqubo zabo ngaphandle kokuyeka ukulawula okunzima komuntuKusazobonakala ukuthi lezi zithembiso zizoshintsha kangakanani zibe izinguquko zangempela ekukhiqizeni kanye nendlela esisebenza ngayo nobuhlakani bokwenziwa ezinyangeni ezizayo.
