I-China Isetha Izinga Lokulawula I-AI Egqamile

I-China Isetha Izinga Lokulawula I-AI Egqamile

I-China Isetha I-Bold AI Regulation Standard ngezinqubomgomo ezintsha ezishanelayo ezilolonga kabusha ingxoxo yomhlaba wonke mayelana nokubusa kobuhlakani bokwenziwa. Idizayinelwe ukuphoqelela ukusebenza obala kwe-algorithmic, izindinganiso zokuziphatha, nokudalwa kokuqukethwe okunesibopho, le mithetho ye-AI ibeka i-China njengesihleli sokulinganiswa somhlaba wonke kuyilapho amazwe aseNtshonalanga ehlala engxoxweni. Njengoba kunemithelela efinyelela kude kubaholi basekhaya bezobuchwepheshe abafana no-Baidu no-Alibaba, kanye nezinkampani zamazwe ngamazwe ezisebenza e-China, lolu hlaka lokulawula aluhlosile nje kuphela ukuvikela ukusetshenziswa kabi kodwa futhi lubonisa isifiso sezwe sokuhola ikusasa lokuphatha ngedijithali.

Okuthathwayo Okubalulekile

• Imithetho ye-AI yase-China igxile ekwazisweni kwe-algorithm, ukusetshenziswa kwe-AI kokuziphatha, nokudalula okuyisibopho kokuqukethwe okukhiqizwa yi-AI.
• Izinkampani zobuchwepheshe kufanele zibike imisebenzi ye-algorithmic futhi zilebule izinto ezikhiqizwe yi-AI ngokusobala.
• Le mithetho isebenza kumafemu aseShayina nezinhlangano zamazwe ngamazwe ezisebenza ngaphakathi kweChina.
• Isimo se-China sokulawula esisebenzayo siphambene nemizamo yokubusa ehamba kancane yase-US kanye ne-EU AI.

I-China's Regulatory Yeap into AI Ethics and Control

Ngomnyakazo onzima, i-China ikhiphe omunye wemithetho ye-AI ebanzi emhlabeni jikelele. Imithetho yenganyelwe i-Cyberspace Administration of China (CAC) futhi ihlanganisa izindawo eziningi, okuhlanganisa ukusetshenziswa kwe-AI okukhiqizayo, ukukhanya kwe-algorithm, ukulebula kokuqukethwe, kanye nokuvimbela ukubandlululwa kwe-algorithmic. Lezi zinyathelo azihlosile kuphela ukuvimbela ukuxhashazwa kwezobuchwepheshe kodwa futhi nokusungula izinhlaka zokuziphendulela zabalingisi bemboni yomphakathi nezimele abasebenzisa izinhlelo ze-AI.

Ezinye zezingxenye eziyinhloko zihlanganisa:

• Ukulebula okuyisibopho kokuqukethwe okudalwe amamodeli e-AI akhiqizayo
• Ukuvinjelwa kokubandlululwa ekuthathweni kwezinqumo kwe-algorithmic
• Ukubika okuyisibopho kokuthi isincomo noma ama-algorithms okukhiqizayo asebenza kanjani
• Izimiso zokusetshenziswa kwezimiso zokuziphatha ezihambisana nezimiso zezenhlalakahle kanye nokuphepha kwezwe

Lesi sinyathelo sihambisana nesu le-China lokuhola umhlaba ekuthuthukisweni kwe-AI kanye nokusatshalaliswa ngenkathi igcina ukulawula okuqinile phezu kwezicelo zayo zomphakathi. Isizwe sesivele sihola ekuqashelweni kobuso kanye nokubhekwa okuhlakaniphile, futhi le mithetho manje ikhulisa lokho kulawula kusikhala se-algorithmic kanye ne-generative AI. Ukuze uthole ukujula kwalokhu kuqhubekela phambili, vakashela isiqephu sendatshana esithi iChina isheshisa phambili emjahweni we-AI.

I-Generative AI Ngaphansi Kokuhlolwa: Izidingo Ezintsha

Amapulatifomu e-Generative AI e-China, kungakhathaliseki ukuthi asekhaya noma angaphandle, abhekana nezingqinamba ezintsha zokusebenza. Lezi zisuka esigabeni sangaphambi kokuqeqeshwa ngokusatshalaliswa komphakathi. Imiyalo ebalulekile ihlanganisa:

• Ukuhlolwa kokuqukethwe: Imiphumela ye-AI akufanele idale, ithuthukise, noma isabalalise amahlebezi, udlame, noma okuqukethwe okubukela phansi ukuzinza komphakathi.
• Ukulebula Okuqukethwe Okuyisibopho: Izinkampani kufanele zibonise isaziso esibonakalayo uma okuqukethwe kukhiqizwa i-AI.
• Ukutholakala Kwedatha Nokuphepha: Idatha yokuqeqeshwa kufanele ithathwe ezintweni ezisemthethweni, ezingabandlululi, nezingephuli umthetho.
• Izindlela Zokubuyekeza Ngaphambili: Wonke amathuluzi e-AI akhiqizwa umphakathi kufanele adlule ekubuyekezweni kobuchwepheshe nokuqukethwe ngaphambi kokwethulwa.

Izinhlawulo zokungathobeli zingabandakanya ukumiswa kwesevisi, izinhlawulo zezezimali, kanye nophenyo lobugebengu ngaphansi kwemithetho yase-China yokuphepha ku-inthanethi nokuphathwa kwedatha. Lokhu kusho ukuthi izinkundla ezifana ne-Google, i-OpenAI, ne-Midjourney manje kufanele zenze amathuluzi e-AI angena ngaphakathi kwalesi sakhiwo esisekelwe emthethweni.

Ngubani Okufanele Athobele: Okuhlosiwe Kwasekhaya Nakwamanye Amazwe

Imithetho iyabophezela kuzo zombili izinkampani zobuchwepheshe zaseShayina kanye namafemu angaphandle asebenza ngaphakathi kwe-ecosystem yedijithali yaseChina. Izinkampani ezinkulu ezifana ne-Tencent, i-Alibaba, i-Huawei, ne-Baidu sezivele zitshale imali kumaphaneli ezimiso ze-AI kanye namathuluzi okuhlola i-algorithm ukuze kuqinisekiswe ukuthotshelwa kwemithetho. Ngisho namafemu omhlaba afana neMicrosoft ne-Apple ayalwa ukuthi abuyekeze izinsiza ezibandakanya izinhlelo zokukhiqiza ze-AI eChina.

Le mithetho ithinta imikhakha efana nenkundla yezokuxhumana, i-e-commerce, ukuqasha, kanye nezinsizakalo zomphakathi. Noma iyiphi inkundla esebenzisa i-AI ukuze ithonye ukuziphatha komsebenzisi noma ikhiqize okuqukethwe kufanele isebenzise amathuluzi okwengamela nokudalula okusobala. Ezinkampanini ezihlose ukuqonda ukubusa okubanzi kwe-AI, indatshana emayelana nezitayela nemithetho ye-AI yokubusa inikeza eminye imininingwane.

I-China vs EU vs US: I-AI Regulation Comparison Table

I-Outlook yochwepheshe: Ukusabela kwamazwe ngamazwe kanye nemithelela

Izazi zomthetho we-AI zigqamisile ukuthi lokhu kugxuma kokulawula kukhombisa ukuzimisela kweChina ukungaholeli nje ekuthuthukisweni kwe-AI kodwa nasekuphatheni. UDkt. Lin Hui waseNyuvesi yaseTsinghua uphawule ngokuthi “Isinyathelo saseChina sokuya phambili mayelana nezimiso zokuziphatha ze-algorithm sizoshintsha indlela izinkampani zomhlaba eziklama ngayo izinhlelo ze-AI ezithobelayo.”

Izinkampani zomhlaba wonke zilungisa izindlela zazo. Umkhulumeli ovela kwa-Google ushilo ngesikhathi sephaneli yakamuva ukuthi le mithetho imele isisekelo esisha sezindawo ezilawulwa kakhulu, futhi ukulinganisa kokuthobela manje sekubaluleke kakhulu.

Abanye abahlaziyi baxwayisa ngokuthi izimfuneko eziqinile zingase zingasheshi ukusungula izinto ezintsha. Iziqalo zobuchwepheshe ezincane nezimaphakathi zingase zibe nzima nezinsiza ezidingekayo ukuze ulawule amathuluzi okuqapha aqinile kanye nokuhlolwa kwamabhuku okuvamile. Naphezu kwalokhu, isiqondiso sicacile. Izinqubomgomo zezwe ze-AI ngokushesha ziba yingxenye yezingxoxo zamasu omhlaba, okuhlanganisa nalezo eziholwa yi-UNESCO kanye nemikhandlu ye-ITU ye-UN.

Umthelela ku-Global AI Governance Trends

Ngenkathi i-US isabumba inqubomgomo yenhlangano ye-AI futhi i-EU iphothula uMthetho wayo we-AI osekelwe engcupheni, umthetho wase-China usuvele uyasebenza. Lokhu kunikeza isizwe ithuba lokuphoqelela futhi kusibeka esikhundleni sokuthumela izindinganiso zaso emhlabeni jikelele. Ukuze uqhathanise nentuthuko yase-US, funda mayelana ne-White House eveza imithetho yokulawula i-AI.

Imibono efana nokuqukethwe kwe-watermarking kwe-AI kanye nokudinga izindawo zokuhlola i-sandbox manje isihamba phambili ngenxa yemizamo yokulawula yase-China yokuqala. Ingxoxo emayelana nokuqondanisa inqubomgomo yomhlaba isiqalile kakade, nezimpawu ezikhomba ekwandeni kwentshisekelo yamazwe ngamazwe ekukopisheni izici zemodeli yaseShayina. Isibonelo, ukudalulwa kwe-algorithm nezibuyekezo zokukhishwa kwangaphambilini kuxoxwa ngazo kaningi njengezici okungaxoxiswana ngazo zezinhlaka zamazwe ngamazwe ezizayo.

Yini Izinkampani Zamazwe Ngamazwe Okufanele Zazi

Kumabhizinisi asebenza e-China, ukuqondanisa amasistimu e-AI nomthetho wendawo akuyona inketho. Ukuze ugweme izinhlawulo noma ukuvinjelwa kwesevisi, izinhlangano kufanele zithathe lezi zinyathelo zokuzilungiselela:

• Hlola izinhlelo ezikhona ze-AI ezisetshenziswa ezimakethe zaseShayina
• Lebula okuqukethwe okukhiqizwa yi-AI kuzo zonke izixhumi ezibonakalayo zabasebenzisi
• Bhala umsebenzi kanye nengqondo yamamodeli e-algorithmic
• Dala amaqembu angaphakathi okugada azinikele kuzimiso zokuziphatha ze-AI nokulawulwa kwezinhlekelele

Ukuthobela imithetho ngokuqhubekayo ngeke nje kunciphise ubungozi. Kungase futhi kusebenze njengenzuzo yokuncintisana njengoba amazwe amaningi eqala ukusebenzisa izindlela ezifanayo zokulawula. Funda kabanzi mayelana nokuthi izinto ezibalulekile ze-AI zase-China zilungisa kanjani kabusha imikhakha kusukela kwezemfundo kuya kubuchwepheshe obukhulu ekusetshenzisweni kwe-AI yase-China emakilasini.

imibuzo ejwayelekile ukubuzwa

Iyini imithetho ye-AI emisha yaseChina?

Imithetho emisha ye-AI yaseChina igunyazwe yiCyberspace Administration yaseChina. Adinga amathuluzi e-AI ukuze alebule ngokucacile okuqukethwe okukhiqizwayo, avimbele izinqubo ze-algorithmic ezibandlululayo, futhi afune ukuba sobala kunengqondo ye-algorithmic. Lokhu kusebenza kuzo zombili izinhlangano zobuchwepheshe zaseShayina nezangaphandle ezisebenza ngaphakathi kweChina.

Unjani umthetho we-AI waseChina uma uqhathaniswa noMthetho we-EU AI?

I-China iphoqelela ukwengamela kwesikhathi sangempela ngokusebenzisa izilawuli ezimaphakathi. Umthetho we-EU AI usebenzisa imodeli yengcuphe enezigaba nokuphoqelela okwabiwe phakathi kwezinhlangano zikazwelonke nezesifunda. Imodeli yaseChina iqinile futhi isetshenziswa ngaphandle kokulinda ukuzilawula kwemboni.

Yimuphi umthetho we-AI iChina esanda kuwumemezela?

I-China imemezele uhlaka oluphelele lwemithetho ye-AI olubeka imihlahlandlela yokuphepha, ubumfihlo bedatha, ukuziphendulela, nokulawulwa kobungozi kuzo zonke izinhlelo ze-AI. Imithetho ihlose ukuphatha ukuthuthukiswa, ukuthunyelwa, kanye nokusetshenziswa kobuhlakani bokwenziwa ukuze kuhambisane nezinto ezibalulekile kuzwelonke kanye nemikhawulo yengozi.

Kungani umthetho we-AI waseChina uthathwa njengesibindi?

Umthetho we-AI wase-China uthathwa njengonesibindi ngoba udlulela ngale kwemithetho yokuphepha emincane ukuze uhlanganise ukuhlolwa kwengozi okuyisibopho, izidingo zokungafihli lutho, ukuziphendulela kwe-algorithmic, kanye nokwengamela umkhakha wonke. Idala izindinganiso ezisebenzisekayo zonjiniyela, ama-deployers, nabanikazi bezinkundla.

Yiziphi izinjongo eziyinhloko zezindinganiso zokulawula i-AI yaseChina?

Izinjongo eziyinhloko ukuqinisekisa ukuphepha kwe-AI, ukuvikela idatha yomuntu siqu, ukuvimbela ukuchema kwe-algorithmic, nokukhuthaza ukuqanjwa okunomthwalo wemfanelo. Amazinga aphinde agxile ekuvikelekeni kwezwe, ukuzinza kwezomnotho, nokuzwana komphakathi ngokuphatha izinhlelo zokusebenza ze-AI ezinobungozi obukhulu.

Yiziphi izinhlangano ezithintwa imithetho emisha ye-AI yaseChina?

Izinkampani zobuchwepheshe, abathuthukisi be-AI, abaqhubi bezinkundla, abahlinzeki besevisi yedatha, nezikhungo zocwaningo ezisebenza ngaphakathi kweChina ziyathinteka. Umthethonqubo udinga ukuthi amabhizinisi enze ukuhlola ubungozi futhi ahambisane nezidingo zokubusa kanye nokubika.

Imithetho ye-AI yase-China ibuthinta kanjani ubumfihlo bedatha?

Umthethonqubo uqinisa ubumfihlo bedatha ngokubeka izidingo eziqinile ekuqoqweni kwedatha, ukugcinwa, ukucutshungulwa, kanye nokusetshenziswa. Amabhizinisi kufanele aqinisekise ukuphathwa kwedatha okusemthethweni, avikele ulwazi lomuntu siqu olubucayi, futhi enze ukuhlola komthelela wobumfihlo.

Ingabe i-China idinga ukucaca kwe-algorithm kumthethonqubo wayo we-AI?

Yebo, ukulawulwa kwe-AI yase-China kudinga ukubonakala kwe-algorithm kumasistimu athinta izintshisekelo zomphakathi noma amalungelo abasebenzisi. Izinhlangano kufanele zinikeze ngamadokhumenti, ukuqonda okuchazekayo, nemikhondo yokuhlola kubalawuli futhi, kwezinye izimo, abasebenzisi.

I-China izisebenzisa kanjani izindinganiso zayo zokulawula i-AI?

I-China isebenzisa izindinganiso ze-AI ngokuhlola, ukucwaninga, ukubika ukuthobela imithetho, kanye nezijeziso zokwephula. Iziphathimandla ezilawulayo zingafuna izenzo zokulungisa, izinhlawulo, noma ukumiswa kwezinsizakalo ze-AI ezibeka ubungozi obungamukeleki.

Ngabe umthetho we-AI waseChina uzozithinta izinkampani zakwamanye amazwe?

Izinkampani zobuchwepheshe zakwamanye amazwe ezisebenza e-China kufanele zithobele izindinganiso zokulawula zemikhiqizo ye-AI namasevisi ahlinzekwa ngaphakathi kwezwe. Lokhu kufaka phakathi ukwenziwa kwendawo kwedatha, ukucwaningwa kokuthotyelwa kwemithetho, nezibopho zokubika ubungozi.

Ingabe zikhona izinhlawulo zokungalandeli imithetho ye-AI yaseChina?

Yebo, ukungathobeli kungabangela izinhlawulo, imikhawulo yokusebenza, ukumiswa kwezinsizakalo, kanye nomthelela esimeni esihle. Abalawuli banegunya lokuphoqelela izinyathelo zokulungisa kanye nezijeziso ezikhulayo zokwephulwa okuphindaphindiwe.

Ngabe umthetho we-AI waseChina uma uqhathaniswa noMthetho we-EU AI?

Umthetho we-AI waseChina kanye noMthetho we-EU AI womabili agcizelela izinhlaka ezisekelwe ebungozini, ukubeka izinto obala, nokuziphendulela. Indlela yaseShayina igxile endaweni eyodwa futhi iqhutshwa uhulumeni, kuyilapho i-EU igxile ekuvumelaniseni izindinganiso kuwo wonke amazwe angamalungu ngokuvikela okuqinile kwamalungelo abantu ngabanye.

Yiziphi izimboni ezithinteka kakhulu yimithetho ye-AI yaseChina?

Izimboni ezithinteke kakhulu zifaka ezezimali, ukunakekelwa kwezempilo, izimoto ezizimele, ukuphepha komphakathi, ezemfundo, kanye nobuchwepheshe babathengi. Noma yimuphi umkhakha osebenzisa amasistimu e-AI athinta ukuphepha, ezezimali, noma idatha yomuntu siqu ungaphansi kokuqondisa okuqinile.

Imithetho ye-AI yaseChina izokwakha kanjani ukubusa kwe-AI emhlabeni jikelele?

Izindinganiso ze-AI zase-China zingaba nomthelela ekubuseni komhlaba wonke ngokubeka izilinganiso zokugada kukahulumeni, ukubonakala kwe-algorithm, nokulawula ubungozi. Ezinye izizwe zingabhekisela noma ziguqule izici ezinhlelweni zazo zokulawula.